framework,version,device,op_name,kernel_source,bmm_dtype,num_tokens,num_heads,latency
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,128,0.006137600168585777
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,128,0.008902399986982345
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,64,0.008191999793052674
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,32,0.0045311998575925825
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,32,0.008339200168848038
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,16,0.004134399816393852
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,16,0.007926400005817413
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,8,0.004099199920892716
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,8,0.007612799853086471
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,4,0.0039007999002933502
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,4,0.007456000149250031
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,64,0.0050592001527547835
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,2,0.003961599990725517
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1,1,0.003824000060558319
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,1,0.0076223999261856076
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,128,0.00626240000128746
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,64,0.005091200023889542
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,8,0.004646399989724159
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,128,0.009251199662685394
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,4,0.0038784001022577285
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,4,0.007552000135183335
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,32,0.004428799822926521
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,2,0.003955199941992759
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,1,0.0038752000778913497
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,1,0.007545600086450577
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2,16,0.004464000090956688
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,128,0.0062304001301527025
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,64,0.0049632001668214795
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,64,0.008819200098514557
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,32,0.004380799829959869
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,32,0.008188799768686295
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,16,0.00421760007739067
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,16,0.00799039974808693
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,8,0.003987200185656547
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,8,0.007654400169849395
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,4,0.003855999931693077
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,4,0.007519999891519547
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,2,0.003903999924659729
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,2,0.007875200361013412
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4,1,0.003936000168323517
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,1,0.007612799853086471
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,128,0.006022400036454201
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,128,0.009904000163078307
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,64,0.004992000013589859
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,64,0.009164799749851228
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,32,0.004387199878692627
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,32,0.008483199775218964
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,16,0.004182400181889534
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,16,0.008028800040483475
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,8,0.003843199834227562
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,8,0.007791999727487564
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,4,0.003916800022125244
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,4,0.007846400141716003
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,2,0.003868800029158592
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,2,0.007612799853086471
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8,1,0.003843199834227562
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8,1,0.007583999633789062
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,128,0.006412799656391144
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1,2,0.00735040009021759
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,64,0.005020799860358238
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,128,0.01156800016760826
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,32,0.004399999976158142
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,32,0.008771199733018875
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,64,0.009814400225877762
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,16,0.004137599840760231
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,8,0.003948799893260002
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,16,0.008198399841785432
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,4,0.004169600084424019
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,8,0.008025600016117096
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,2,0.0038943998515605927
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,4,0.007977599650621415
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,16,1,0.004220800101757049
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,2,0.007791999727487564
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,16,1,0.007731200009584427
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,128,0.006649599969387054
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,64,0.005302400141954422
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,128,0.015244799852371215
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,64,0.011459200084209442
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,32,0.004595199972391129
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,32,0.009350399672985076
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,16,0.004284799844026565
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,16,0.008713600039482117
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,8,0.004179200157523155
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,4,0.004016000032424927
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,8,0.008124800026416778
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,2,0.0042975999414920805
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,4,0.00799039974808693
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,32,1,0.003987200185656547
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,2,0.007913599908351897
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,32,1,0.00772159993648529
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,128,0.007174400240182876
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,64,0.005548800155520439
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,128,0.018934400379657747
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,32,0.004748800024390221
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,64,0.013558399677276612
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,32,0.010233599692583084
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,16,0.004464000090956688
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,8,0.004249599948525429
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,16,0.009017600119113922
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,4,0.004095999896526337
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,8,0.008217599987983704
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,4,0.008198399841785432
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,2,0.004143999889492989
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,2,0.007955200225114822
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,48,1,0.004086399823427201
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,48,1,0.0077855996787548065
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,128,0.007804799824953079
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,128,0.022806400060653688
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,64,0.00599679984152317
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,64,0.015219199657440185
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,32,0.004812800139188766
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,32,0.011177600175142289
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,16,0.0042975999414920805
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,32,0.007865600287914276
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,64,0.008403199911117553
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,16,0.009174399822950364
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,4,0.004121600091457367
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,8,0.00822720006108284
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,4,0.008371199667453765
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,16,0.007580800354480744
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,2,0.007936000078916549
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,64,1,0.00799039974808693
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,128,0.008406399935483932
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,128,0.026060798764228822
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,64,0.0063231997191905975
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,64,0.016995200514793397
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,32,0.005033599957823753
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,32,0.012147200107574464
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,16,0.0047136001288890835
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,8,0.0073183998465538025
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2,2,0.007366400212049484
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,16,0.009855999797582626
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4,128,0.009286399930715561
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,8,0.004278400167822838
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,4,0.004214400053024292
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,8,0.008115199953317642
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,2,0.0041184000670909885
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,4,0.007760000228881836
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,2,0.008310399949550629
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,8,0.004307200014591217
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,80,1,0.004163200035691261
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,80,1,0.008188799768686295
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,128,0.009142400324344635
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,128,0.02985920011997223
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,64,0.00650240033864975
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,2,0.00416640006005764
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,64,0.018486399948596955
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,32,0.005385600030422211
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,16,0.004774399846792221
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,32,0.012969599664211273
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,4,0.004172800108790398
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,64,1,0.003999999910593033
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,8,0.004233599826693535
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,8,0.008892799913883209
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,16,0.010175999999046326
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,1,0.0038975998759269714
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,4,0.008428800106048583
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,2,0.008207999914884568
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,96,2,0.005462399870157242
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,96,1,0.008207999914884568
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,128,0.011151999980211259
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,128,0.0377375990152359
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,64,0.007747200131416321
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,32,0.005635200068354607
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,64,0.02258239984512329
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,8,0.004252799972891808
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,32,0.014921599626541137
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,16,0.004665600135922432
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,16,0.011046399921178817
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,4,0.004243199899792671
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,2,0.004080000147223472
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,4,0.008495999872684479
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,8,0.009142400324344635
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,2,0.008111999928951263
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,128,1,0.008182399719953538
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,128,1,0.004095999896526337
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,128,0.01583999991416931
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,64,0.008284799754619598
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,128,0.0460319995880127
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,64,0.026137599349021913
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,32,0.016787199676036833
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,32,0.005958399921655655
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,16,0.004704000055789947
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,8,0.004371200129389763
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,16,0.012124799937009812
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,2,0.004176000133156776
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,4,0.008752000331878663
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,8,0.00976639986038208
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,2,0.008252800256013871
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,4,0.004195199906826019
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,160,1,0.00811839997768402
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,160,1,0.004128000140190125
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,128,0.02197439968585968
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,32,0.006492800265550614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,128,0.057068800926208495
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,32,0.018486399948596955
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,64,0.02963840067386627
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,16,0.004796800017356872
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,64,0.008774399757385254
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,8,0.010060799866914749
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,8,0.0046720001846551895
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,16,0.01311360001564026
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,4,0.00952960029244423
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,4,0.004249599948525429
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,2,0.004006399959325791
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,2,0.008460800349712371
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,192,1,0.008432000130414962
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,192,1,0.004016000032424927
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,128,0.026473599672317504
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,16,0.005584000051021576
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,32,0.007187200337648391
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,128,0.07128319740295411
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,32,0.021964800357818604
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,64,0.03650240004062653
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,64,0.009987200051546097
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,8,0.004652800038456917
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,16,0.014975999295711518
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,4,0.00424639992415905
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,8,0.011023999750614166
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,1,0.004009599983692169
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,4,0.00896959975361824
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,256,2,0.004121600091457367
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,1,0.008297599852085114
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,256,2,0.008364800363779068
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,64,0.01284479945898056
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,128,0.029600000381469725
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,64,0.04658240079879761
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,32,0.007462400197982788
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,32,0.025836798548698425
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,16,0.005936000123620033
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,8,0.004783999919891357
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,128,0.08671680092811584
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,16,0.01650879979133606
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,8,0.012278400361537933
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,4,0.004969599843025208
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,2,0.004227200150489807
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,4,0.00986879989504814
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,2,0.008508799970149994
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,320,1,0.00811839997768402
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,320,1,0.004239999875426293
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,64,0.015904000401496886
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,128,0.10235840082168579
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,64,0.05574079751968384
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,128,0.034544000029563905
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,16,0.006335999816656113
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,32,0.029651200771331786
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,8,0.0048640001565217975
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,16,0.018361599743366243
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,32,0.008515200018882752
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,4,0.004566400125622749
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,8,0.013027200102806091
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,2,0.004320000112056732
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,4,0.009836799651384353
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,2,0.008806400001049042
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,384,1,0.003971200063824653
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,384,1,0.008294399827718735
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,128,0.04182719886302948
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,64,0.02266560047864914
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,16,0.0071712002158164975
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,64,0.07111039757728577
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,128,0.13309119939804076
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,32,0.010000000149011612
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,4,0.004614400118589402
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,32,0.03686720132827759
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,8,0.005311999842524529
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,4,0.010995200276374817
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,16,0.02210559993982315
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,8,0.014697599411010741
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,1,0.008380799740552902
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,1,0.0040608000010252
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,512,2,0.009251199662685394
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,512,2,0.004249599948525429
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,64,0.1023360013961792
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,128,0.05727360248565674
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,64,0.03118720054626465
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,128,0.19498879909515382
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,32,0.01438400000333786
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,16,0.029452800750732422
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,32,0.0551584005355835
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,8,0.018409599363803864
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,8,0.006335999816656113
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,4,0.005023999884724617
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,2,0.010054399818181991
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,2,0.004348799958825112
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,16,0.008556800335645676
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,4,0.012854400277137756
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,768,1,0.0042015999555587765
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,128,0.0722208023071289
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,768,1,0.00896959975361824
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,64,0.03956480026245117
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,64,0.13222399950027466
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,16,0.009990400075912476
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,8,0.007094399631023407
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,32,0.021955199539661407
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,16,0.03700799942016601
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,128,0.25943679809570314
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,32,0.07083839774131775
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,4,0.005289600044488907
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,8,0.02215680032968521
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,4,0.014787200093269347
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,2,0.004726399853825569
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,2,0.010841599851846694
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1024,1,0.0043935999274253845
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1024,1,0.00920960009098053
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,128,0.10289599895477294
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,32,0.03028160035610199
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,64,0.05464000105857849
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,32,0.10216319561004639
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,64,0.1959231972694397
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,16,0.014211200177669525
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,128,0.3860320091247559
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,8,0.00846719965338707
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,4,0.00628800019621849
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,16,0.05478720068931579
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,2,0.004899200052022934
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,8,0.029718399047851562
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,4,0.018479999899864197
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,2,0.012886400520801543
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,1536,1,0.0044511999934911724
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,1536,1,0.010016000270843506
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,128,0.1340000033378601
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,64,0.0700160026550293
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,32,0.03818239867687225
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,16,0.02006080001592636
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,32,0.13233920335769653
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,8,0.009855999797582626
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,16,0.07037439942359924
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,128,0.5135712146759033
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,64,0.25744640827178955
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,4,0.00719040036201477
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,8,0.03702079951763153
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,4,0.021958400309085847
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,1,0.004508800059556961
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,2048,2,0.005392000079154968
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,1,0.010896000266075134
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,2048,2,0.014905600249767304
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,128,0.19534720182418824
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,64,0.1010975956916809
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,32,0.053465598821640016
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,16,0.029731199145317078
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,8,0.014364799857139588
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,32,0.19488320350646973
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,16,0.10191359519958496
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,64,0.38391358852386476
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,8,0.055452799797058104
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,4,0.029241600632667543
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,2,0.006412799656391144
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,128,0.7750495910644531
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,4,0.00841279998421669
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,3072,1,0.004956800118088722
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,2,0.01870400011539459
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,3072,1,0.013020800054073333
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,64,0.13195199966430665
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,128,0.25711679458618164
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,32,0.06875200271606445
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,16,0.03749119937419891
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,16,0.13201279640197755
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,8,0.020134399831295013
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,32,0.2571104049682617
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,64,0.5116159915924072
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,4,0.010016000270843506
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,8,0.07053120136260986
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,4,0.036617600917816163
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,2,0.007184000313282013
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,2,0.021996800601482392
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,1,0.014720000326633453
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,4096,1,0.005465599894523621
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,4096,128,1.0354911804199218
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,64,0.19323840141296386
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,128,0.3810496091842651
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,32,0.09933120012283325
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,16,0.0534496009349823
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,8,0.029151999950408937
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,32,0.3821664094924927
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,4,0.014265599846839904
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,8,0.10151679515838623
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,16,0.19512959718704223
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,2,0.02956480085849762
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,64,0.7725120067596436
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,4,0.05464959740638733
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,2,0.008640000224113464
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,6144,1,0.006227200105786324
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,1,0.01830720007419586
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,32,0.1301535964012146
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,64,0.25443520545959475
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,6144,128,1.561526393890381
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,16,0.0682752013206482
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,128,0.504915189743042
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,8,0.037360000610351565
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,4,0.02011519968509674
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,16,0.2570591926574707
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,32,0.5102208137512207
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,2,0.00979200005531311
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,64,1.0390591621398926
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,8,0.1319648027420044
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,float16,8192,1,0.007056000083684922
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,4,0.07022719979286193
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,2,0.03666560053825378
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,1,0.022291199862957002
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_pre,default,fp8,8192,128,2.086182403564453
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,128,0.006006399914622307
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,128,0.008767999708652496
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,64,0.005110400170087815
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,64,0.008694399893283845
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,32,0.004927999898791313
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,32,0.008326400071382523
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,16,0.00451200008392334
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,16,0.008003199845552445
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,8,0.005161599814891815
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,8,0.00992320030927658
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,4,0.004521600157022476
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,2,0.004390399903059006
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,2,0.007791999727487564
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1,1,0.004396799951791763
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,1,0.007740800082683563
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,128,0.005974400043487549
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,128,0.009113600105047226
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,64,0.005008000135421753
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,64,0.008454400300979614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,32,0.008617600053548813
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,16,0.004553600028157234
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,16,0.008160000294446945
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,8,0.004345599934458733
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,4,0.004534399881958961
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,4,0.007999999821186066
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,2,0.004761600121855736
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,2,0.00793280005455017
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,1,0.004447999969124794
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,128,0.009507200121879578
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,64,0.005129599943757057
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,64,0.008819200098514557
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,32,0.008483199775218964
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,16,0.00822720006108284
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,8,0.004403200000524521
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,8,0.008051200211048127
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,4,0.0045311998575925825
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,4,0.008012799918651581
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,2,0.004345599934458733
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,2,0.007971200346946716
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,1,0.004387199878692627
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4,1,0.008060800284147263
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2,32,0.004633599892258644
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,128,0.0058559998869895935
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,64,0.004755200073122978
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1,4,0.00748480036854744
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,128,0.005846399813890457
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,64,0.008934400230646133
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,32,0.008483199775218964
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,16,0.004323200136423111
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,16,0.008310399949550629
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,8,0.004329600185155868
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,4,0.004588799923658371
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,8,0.008217599987983704
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,2,0.004399999976158142
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,2,0.008051200211048127
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,1,0.004371200129389763
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,1,0.008099199831485748
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,128,0.0061055999249219894
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,16,0.004921599850058555
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4,32,0.004700800031423568
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,64,0.0047520000487566
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,32,0.004518400132656098
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,128,0.011472000181674958
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8,32,0.004483199864625931
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,16,0.004396799951791763
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,16,0.008476799726486206
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,8,0.004291199892759323
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,8,0.00823040008544922
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,4,0.004441599920392036
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,4,0.00822720006108284
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,2,0.004844800010323525
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,2,0.007936000078916549
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16,1,0.004275200143456459
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,128,0.006284800171852112
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,1,0.00777600035071373
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,128,0.01499519944190979
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,64,0.0057888001203536986
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,32,0.0047520000487566
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,64,0.011260800063610077
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,16,0.004639999940991402
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,32,0.009507200121879578
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,16,0.00875839963555336
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,8,0.004428799822926521
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,8,0.00841279998421669
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,4,0.004467200115323066
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,4,0.008348800241947174
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,2,0.004569600149989128
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,2,0.008220800012350083
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,32,1,0.004335999861359597
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,32,1,0.00809599980711937
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,128,0.006761600077152252
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,128,0.018479999899864197
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,64,0.0051968000829219815
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,32,0.004828799888491631
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,64,0.012860800325870513
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,16,0.004729599878191948
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,32,0.010060799866914749
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,16,0.00912960022687912
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,8,0.004566400125622749
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,8,0.008668799698352814
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,4,0.00453759990632534
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,4,0.008499199897050858
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,2,0.004467200115323066
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,2,0.008352000266313553
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,48,1,0.00451200008392334
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,48,1,0.008457600325345992
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,8,0.007456000149250031
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,128,0.007286400347948074
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,64,0.00562559999525547
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,128,0.02096959948539734
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2,1,0.007353600114583969
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,64,0.014678399264812469
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,16,0.004800000041723251
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,32,0.010547199845314026
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,16,0.009471999853849411
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,8,0.008841600269079208
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,32,0.0046016000211238865
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,4,0.004556800052523613
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,2,0.004464000090956688
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,4,0.008479999750852585
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,128,0.009305600076913834
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8,4,0.00806720033288002
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,2,0.00833280012011528
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,1,0.004447999969124794
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,128,0.007820799946784973
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,64,1,0.007737600058317184
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,32,0.004735999926924706
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,64,8,0.0045311998575925825
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,128,0.024905599653720856
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,64,0.005660799890756607
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,16,0.004735999926924706
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,64,0.016371199488639833
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,32,0.0121568001806736
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,8,0.004604800045490265
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,4,0.004521600157022476
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,8,0.0091839998960495
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,4,0.008323200047016144
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,64,0.008899199962615966
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,1,0.004582399874925614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16,32,0.008857599645853042
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,80,2,0.004534399881958961
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,16,0.009945599734783173
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,2,0.008473599702119828
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,80,1,0.008454400300979614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,128,0.00801599994301796
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,128,0.027958399057388304
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,64,0.005872000008821487
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,32,0.005071999877691269
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,64,0.0178399994969368
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,16,0.004758400097489357
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,16,0.010051199793815612
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,32,0.012963199615478515
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,8,0.004678399860858917
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,4,0.008736000210046769
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,4,0.004499199986457825
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,2,0.008457600325345992
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,1,0.00823040008544922
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,2,0.004403200000524521
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,96,8,0.009087999910116195
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,96,1,0.004329600185155868
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,64,0.0063391998410224915
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,32,0.005296000093221664
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,64,0.021142399311065672
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,16,0.0048096001148223875
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,128,0.03792960047721863
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,128,0.008966399729251862
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,8,0.004649600014090538
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,32,0.014412799477577209
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,16,0.010956799983978272
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,4,0.004595199972391129
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,8,0.009507200121879578
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,2,0.004473600164055824
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,4,0.008604799956083297
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,2,0.008454400300979614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,128,1,0.004447999969124794
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,128,1,0.008399999886751174
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,128,0.017231999337673186
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,64,0.024582399427890776
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,32,0.016067199409008026
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,128,0.04630399942398071
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,64,0.006911999732255936
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,16,0.004995200037956238
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,8,0.0048191998153924945
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,16,0.012243200093507767
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,8,0.009756799787282944
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,32,0.005404800176620483
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,4,0.008947200328111648
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,1,0.004460800066590309
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,1,0.008396799862384795
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,4,0.00469760000705719
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,160,2,0.006527999788522721
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,128,0.02096640020608902
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,160,2,0.008550400286912918
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,32,0.005657599866390228
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,32,0.01781439930200577
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,64,0.0071103997528553006
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,16,0.005110400170087815
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,64,0.02805120050907135
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,2,0.0084927998483181
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,2,0.0045471999794244765
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,4,0.004572800174355507
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,128,0.05434240102767944
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,1,0.008316799998283386
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,8,0.005071999877691269
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,4,0.009094399958848953
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,16,0.012966400384902954
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,192,1,0.004371200129389763
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,192,8,0.010543999820947647
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,64,0.007894399762153625
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,128,0.024502399563789367
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,32,0.005974400043487549
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,32,0.021055999398231506
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,64,0.036284801363945005
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,16,0.005030399933457374
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,128,0.06949120163917541
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,4,0.004604800045490265
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,2,0.004524800181388855
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,16,0.014486399292945863
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,8,0.01080000028014183
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,1,0.008380799740552902
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,8,0.004902400076389313
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,4,0.009494400024414063
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,256,1,0.004588799923658371
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,256,2,0.008937600255012512
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,128,0.02807359993457794
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,64,0.0088639996945858
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,128,0.08521919846534728
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,32,0.006764800101518631
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,64,0.044896000623703004
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,16,0.01592320054769516
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,8,0.005267199873924255
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,32,0.024700799584388734
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,4,0.004726399853825569
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,16,0.005158400163054466
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,1,0.00838399976491928
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,8,0.012332800030708312
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,2,0.008851200342178345
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,2,0.004508800059556961
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,320,1,0.004614400118589402
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,320,4,0.00974079966545105
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,16,0.005475199967622757
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,64,0.01281599998474121
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,128,0.03226560056209564
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,64,0.052825599908828735
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,16,0.01783040016889572
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,32,0.007081600278615952
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,8,0.004975999891757965
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,4,0.004796800017356872
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,32,0.02773120105266571
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,1,0.004636799916625023
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,128,0.1025696039199829
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,1,0.008483199775218964
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,4,0.010236799716949463
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,2,0.009164799749851228
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,384,2,0.004665600135922432
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,384,8,0.012921600043773651
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,128,0.03909119963645935
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,64,0.022204799950122832
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,32,0.007999999821186066
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,128,0.1335360050201416
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,64,0.06872959733009339
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,32,0.03505280017852783
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,8,0.0050432000309228895
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,16,0.020819200575351714
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,16,0.005744000151753426
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,4,0.004806400090456009
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,8,0.014486399292945863
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,1,0.004438399896025657
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,512,2,0.0047391999512910845
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,2,0.009308800101280212
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,4,0.011126399785280228
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,128,0.054502397775650024
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,64,0.029180800914764403
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,512,1,0.0088128000497818
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,64,0.10095360279083251
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,32,0.009904000163078307
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,16,0.02770879864692688
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,16,0.007225599884986877
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,8,0.01780479997396469
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,8,0.005379199981689453
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,4,0.005084799975156784
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,32,0.052300798892974856
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,2,0.0048767998814582825
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,4,0.012851199507713318
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,128,0.19516479969024658
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,2,0.010367999970912933
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,768,1,0.004582399874925614
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,768,1,0.009084799885749817
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,128,0.06975679993629455
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,64,0.03676480054855347
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,32,0.01923840045928955
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,128,0.2574656009674072
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,8,0.005699200183153152
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,16,0.00806720033288002
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,64,0.1317247986793518
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,16,0.03499200046062469
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,32,0.06845440268516541
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,4,0.0051392000168561935
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,8,0.02184319943189621
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,1,0.004483199864625931
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,2,0.011110399663448334
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,4,0.014588800072669984
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1024,2,0.004822399839758873
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1024,1,0.00947519987821579
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,128,0.09802560210227966
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,32,0.02812480032444
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,64,0.05303360223770141
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,16,0.010025600343942643
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,8,0.007283200323581695
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,64,0.19561280012130738
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,16,0.052172797918319705
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,128,0.3843679904937744
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,2,0.004924799874424934
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,32,0.10062400102615357
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,4,0.005407999828457832
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,4,0.01780479997396469
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,8,0.02797439992427826
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,1536,1,0.0047680001705884935
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,1,0.010047999769449234
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,1536,2,0.013040000200271606
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,64,0.0666208028793335
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,32,0.0355679988861084
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,128,0.1271232008934021
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,16,0.018169599771499633
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,64,0.2562560081481934
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,32,0.1310271978378296
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,16,0.06811519861221313
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,8,0.035359999537467955
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,2,0.004950400069355964
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,4,0.005724800005555153
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,4,0.020950399339199066
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,8,0.007987199723720551
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,128,0.5227456092834473
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,2048,1,0.004790399968624115
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,1,0.011126399785280228
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,2048,2,0.014399999380111694
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,128,0.18498879671096802
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,32,0.0512287974357605
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,64,0.09473919868469238
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,32,0.19434560537338258
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,16,0.02999039888381958
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,64,0.3840224027633667
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,8,0.009769599884748459
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,16,0.09955840110778809
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,8,0.052153599262237546
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,4,0.027955201268196107
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,2,0.005433600023388862
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,4,0.007331199944019318
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,128,0.795740795135498
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,3072,1,0.005008000135421753
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,2,0.01770240068435669
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,3072,1,0.013027200102806091
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,32,0.06505600214004517
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,128,0.25122880935668945
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,64,0.12261120080947877
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,16,0.03497599959373474
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,8,0.01698880046606064
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,32,0.2558624029159546
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,4,0.00801599994301796
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,2,0.005644800141453743
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,8,0.06761919856071472
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,16,0.13079999685287474
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,64,0.5149087905883789
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,4096,1,0.004969599843025208
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,2,0.02099200040102005
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,4,0.03513599932193756
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,1,0.014547200500965118
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,4096,128,1.0997471809387207
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,64,0.17887040376663207
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,32,0.09436479806900025
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,16,0.05643200278282166
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,128,0.3717600107192993
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,8,0.026675200462341307
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,16,0.19378559589385985
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,32,0.3840831995010376
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,4,0.009609600156545639
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,64,0.8057920455932617
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,4,0.052262401580810545
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,8,0.09932799935340882
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,2,0.007081600278615952
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,6144,1,0.005171199887990951
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,2,0.027939200401306152
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,1,0.01783040016889572
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,32,0.12138240337371826
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,64,0.2453887939453125
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,6144,128,1.6674079895019531
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,128,0.4693439960479736
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,16,0.06479039788246155
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,8,0.035257598757743834
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,4,0.016361600160598753
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,16,0.25583999156951903
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,2,0.008019199967384339
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,2,0.034857600927352905
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,4,0.06772480010986329
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,64,1.075331211090088
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,8,0.13626240491867064
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,32,0.518287992477417
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,8192,1,0.00565119981765747
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,1,0.020947200059890748
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,32,0.17819839715957642
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,64,0.34454400539398194
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,128,0.7020031929016113
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,16,0.10563199520111084
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,32,0.8081760406494141
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,8192,128,2.2488096237182615
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,8,0.05030080080032349
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,4,0.029209598898887634
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,2,0.00971520021557808
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,8,0.19481279850006103
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,4,0.09897599816322326
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,12288,1,0.0072223998606204985
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,16,0.39567999839782714
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,2,0.05240960121154785
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,64,1.6650880813598632
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,1,0.027804800868034364
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,32,0.2331167936325073
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,64,0.4573631763458252
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,16,0.1411679983139038
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,128,0.9311103820800781
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,8,0.06650879979133606
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,16,0.5129312038421631
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,12288,128,3.401580810546875
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,32,1.0936256408691407
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,4,0.033904001116752625
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,2,0.017577600479125977
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,8,0.25893120765686034
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,2,0.06781119704246522
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,1,0.03577919900417328
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,4,0.13634560108184815
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,16384,1,0.007740800082683563
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,64,2.2370784759521483
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,64,0.5682975769042968
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,32,0.33314239978790283
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,16,0.1543455958366394
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,128,1.1585536003112793
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,8,0.08121280074119568
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,16,0.660748815536499
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,16384,128,4.502880096435547
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,4,0.04166080057621002
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,8,0.3213664054870605
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,32,1.386134433746338
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,2,0.022575999796390533
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,2,0.08271999955177307
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,float16,20480,1,0.008505599945783615
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,64,2.720425605773926
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,4,0.1700063943862915
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,1,0.04484800100326538
SGLang,0.5.8.post1,NVIDIA H100 80GB HBM3,mla_gen_post,default,fp8,20480,128,5.67454719543457
