Xeon E3-1245 v5 Ubuntu Linux 5.4, "RealSR-NCNN 20200818 - Scale: 4x - TAA: No", Lower Results Are Better "Linux 5.4",17.965,16.894,17.038,17.065 "Linux 5.8",283.647,283.379,283.475 "Linux 5.9 Git",283.426,283.462,283.499 "NCNN 20200916 - Target: Vulkan GPU - Model: mobilenet", Lower Results Are Better "Linux 5.4",45.95,46.05,47.3 "Linux 5.8",38.3,39.1,38.38 "Linux 5.9 Git",38.35,38.45,37.7 "NCNN 20200916 - Target: Vulkan GPU - Model: squeezenet", Lower Results Are Better "Linux 5.4",52.12,52.12,52.14 "Linux 5.8",43.82,43.78,43.8 "Linux 5.9 Git",43.78,43.78,43.84 "NCNN 20200916 - Target: Vulkan GPU - Model: googlenet", Lower Results Are Better "Linux 5.4",42.18,42.16,42.14 "Linux 5.8",35.49,35.46,35.48 "Linux 5.9 Git",35.44,35.49,35.47 "NCNN 20200916 - Target: Vulkan GPU - Model: efficientnet-b0", Lower Results Are Better "Linux 5.4",30.9,30.84,30.85 "Linux 5.8",26.03,26.03,26 "Linux 5.9 Git",26.02,26.01,26.03 "LeelaChessZero 0.26 - Backend: BLAS", Higher Results Are Better "Linux 5.4",796,796,798 "Linux 5.8",679,662,694 "Linux 5.9 Git",777,777,784 "NCNN 20200916 - Target: Vulkan GPU - Model: resnet18", Lower Results Are Better "Linux 5.4",36.81,36.79,36.8 "Linux 5.8",31.34,31.33,31.34 "Linux 5.9 Git",31.38,31.38,31.33 "NCNN 20200916 - Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3", Lower Results Are Better "Linux 5.4",16.34,16.3,16.31 "Linux 5.8", "Linux 5.9 Git",13.99,14,13.99 "NCNN 20200916 - Target: Vulkan GPU - Model: alexnet", Lower Results Are Better "Linux 5.4",58.04,57.82,57.97 "Linux 5.8",51.13,51.48,51.28 "Linux 5.9 Git",50.37,48.05,50.92 "NCNN 20200916 - Target: Vulkan GPU - Model: mnasnet", Lower Results Are Better "Linux 5.4",15.13,15.09,15.11 "Linux 5.8", "Linux 5.9 Git", "NCNN 20200916 - Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2", Lower Results Are Better "Linux 5.4",14.71,14.7,14.75 "Linux 5.8",12.67,12.69,12.7 "Linux 5.9 Git",12.68,12.69,12.67 "NCNN 20200916 - Target: Vulkan GPU - Model: resnet50", Lower Results Are Better "Linux 5.4",87.1,87.11,87.11 "Linux 5.8",75.27,75.23,75.26 "Linux 5.9 Git",75.26,75.28,75.28 "GLmark2 2020.04 - Resolution: 3840 x 2160", Higher Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "NCNN 20200916 - Target: Vulkan GPU - Model: yolov4-tiny", Lower Results Are Better "Linux 5.4",89.38,89.24,89.39 "Linux 5.8",79.42,79.49,79.45 "Linux 5.9 Git",79.31,79.38,79.29 "Apache CouchDB 3.1.1 - Bulk Size: 100 - Inserts: 1000 - Rounds: 24", Lower Results Are Better "Linux 5.4",178.475,174.354,174.544 "Linux 5.8",164.322,154.517,155.135,155.239 "Linux 5.9 Git",166.358,153.735,161.051,164.029,162.197 "GLmark2 2020.04 - Resolution: 1920 x 1080", Higher Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "oneDNN 1.5 - Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",286.701,286.555,284.627 "Linux 5.8",275.169,272.884,263.942 "Linux 5.9 Git",248.577,267.197,270.117,252.591,270.034,269.066,252.296,272.637,268.409,253.987,268.847,252.108,249.768,251.656,267.165 "oneDNN 1.5 - Harness: Deconvolution Batch deconv_1d - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",9.25074,9.25096,9.27341 "Linux 5.8",9.91397,9.66238,9.63102 "Linux 5.9 Git",9.63277,9.84911,10.0017 "oneDNN 1.5 - Harness: Deconvolution Batch deconv_3d - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",14.3001,14.4109,14.3842 "Linux 5.8",15.1342,15.099,15.2017 "Linux 5.9 Git",14.6696,14.8539,14.6216 "NCNN 20200916 - Target: Vulkan GPU - Model: vgg16", Lower Results Are Better "Linux 5.4",214.17,214.11,214.36 "Linux 5.8",206.46,206.63,206.35 "Linux 5.9 Git",204.82,205.56,205.07 "Mlpack Benchmark - Benchmark: scikit_ica", Lower Results Are Better "Linux 5.4",70.373179674149,70.644183158875,72.28666973114 "Linux 5.8",67.125159978867,67.326308012009,69.948958396912 "Linux 5.9 Git",70.212180376053,70.705380678177,69.249639749527 "KeyDB 6.0.16 - ", Higher Results Are Better "Linux 5.4",395120.97,392648.16,393039.88 "Linux 5.8",381604.89,381639.66,380459.4 "Linux 5.9 Git",377496.4,378271.76,377042.03 "oneDNN 1.5 - Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",569.18,569.32,569.691 "Linux 5.8",546.589,548.282,549.26 "Linux 5.9 Git",552.76,560.375,563.455 "Crafty 25.2 - Elapsed Time", Higher Results Are Better "Linux 5.4",7520073,7505088,7616799 "Linux 5.8",7511330,7467956,7428608 "Linux 5.9 Git",7267436,7297323,7342009 "oneDNN 1.5 - Harness: IP Batch 1D - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",3.76718,3.78027,3.79244 "Linux 5.8",3.89922,3.9007,3.90539 "Linux 5.9 Git",3.85496,3.8356,3.85091 "oneDNN 1.5 - Harness: IP Batch All - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",110.818,110.828,111.699 "Linux 5.8",117.115,113.612,111.982 "Linux 5.9 Git",110.577,110.652,110.918 "oneDNN 1.5 - Harness: IP Batch 1D - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",8.18121,8.17315,8.16264 "Linux 5.8",8.37945,8.42913,8.3226 "Linux 5.9 Git",8.389,8.35254,8.47954 "DeepSpeech 0.6 - Acceleration: CPU", Lower Results Are Better "Linux 5.4",78.51346,78.42386,78.6026 "Linux 5.8",81.62231,80.10061,80.10572 "Linux 5.9 Git",81.10888,79.50029,79.53329 "LeelaChessZero 0.26 - Backend: Eigen", Higher Results Are Better "Linux 5.4",720,724,706 "Linux 5.8",622,694,710,701,725,722,696,717,718 "Linux 5.9 Git",724,713,721 "oneDNN 1.5 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",5.14145,5.16707,5.16392 "Linux 5.8",5.01914,5.0619,5.06835 "Linux 5.9 Git",5.02718,5.05119,5.05639 "oneDNN 1.5 - Harness: IP Batch All - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",49.8633,49.849,49.8705 "Linux 5.8",50.9956,51.2528,50.6651 "Linux 5.9 Git",50.2325,50.5182,49.9833 "System GZIP Decompression - ", Lower Results Are Better "Linux 5.4",3.282,3.284,3.281 "Linux 5.8",3.793,3.31,3.298,3.307,3.3,3.293,3.301,3.301,3.299,3.306,3.294,3.3,3.299,3.339 "Linux 5.9 Git",3.903,3.306,3.295,3.335,3.333,3.297,3.296,3.332,3.291,3.334,3.298,3.299,3.298 "Mlpack Benchmark - Benchmark: scikit_linearridgeregression", Lower Results Are Better "Linux 5.4",4.3778140544891,4.2283899784088,4.2710580825806 "Linux 5.8",4.2450306415558,4.2810900211334,4.196001291275 "Linux 5.9 Git",4.2544240951538,4.4445781707764,4.19810795784,4.422527551651 "Timed GDB GNU Debugger Compilation 9.1 - Time To Compile", Lower Results Are Better "Linux 5.4",159.923,159.238,159.65 "Linux 5.8",157.813,156.998,157.159 "Linux 5.9 Git",161.084,160.064,160.761 "oneDNN 1.5 - Harness: Deconvolution Batch deconv_1d - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",10.4834,10.4573,10.2434 "Linux 5.8",10.5254,10.2337,10.163 "Linux 5.9 Git",10.163,10.1876,10.2045 "NCNN 20200916 - Target: CPU - Model: alexnet", Lower Results Are Better "Linux 5.4",22.11,22.82,22.74 "Linux 5.8",22.1,22.14,22.12 "Linux 5.9 Git",22.11,22.15,22.97 "eSpeak-NG Speech Engine 20200907 - Text-To-Speech Synthesis", Lower Results Are Better "Linux 5.4",30.613,34.22,34.118,34.134,34.225,34.242,34.131,34.121,34.078,33.95,34.222,34.068,34.171,34.176,34.22,34.243 "Linux 5.8",33.684,34.568,34.388,34.389 "Linux 5.9 Git",34.368,34.956,34.548,34.481 "VkFFT 2020-09-29 - ", Higher Results Are Better "Linux 5.4",1140,1142,1142 "Linux 5.8", "Linux 5.9 Git", "Montage Astronomical Image Mosaic Engine 6.0 - Mosaic of M17, K band, 1.5 deg x 1.5 deg", Lower Results Are Better "Linux 5.4",89.46,89.609,89.609 "Linux 5.8",90.714,90.544,90.832 "Linux 5.9 Git",91.297,91.119,90.682 "Hugin - Panorama Photo Assistant + Stitching Time", Lower Results Are Better "Linux 5.4",74.406,73.183,72.801 "Linux 5.8",74.507,74.021,73.639 "Linux 5.9 Git",75.325,74.749,73.877 "Rodinia 3.1 - Test: OpenMP Leukocyte", Lower Results Are Better "Linux 5.4",255.266,254.022,254.737 "Linux 5.8",256.259,257.19,257.3 "Linux 5.9 Git",261.074,257.747,257.496 "Mobile Neural Network 2020-09-17 - Model: resnet-v2-50", Lower Results Are Better "Linux 5.4",56.074,55.939,56.48 "Linux 5.8",57.192,56.842,56.767 "Linux 5.9 Git",57.164,57.048,56.92 "libavif avifenc 0.7.3 - Encoder Speed: 8", Lower Results Are Better "Linux 5.4",8.187,8.184,8.225 "Linux 5.8",8.301,8.229,8.242 "Linux 5.9 Git",8.305,8.315,8.329 "NCNN 20200916 - Target: CPU - Model: blazeface", Lower Results Are Better "Linux 5.4",2.14,2.21,2.15 "Linux 5.8",2.14,2.14,2.15 "Linux 5.9 Git",2.15,2.15,2.14 "ECP-CANDLE 0.3 - Benchmark: P3B2", Lower Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "Mobile Neural Network 2020-09-17 - Model: SqueezeNetV1.0", Lower Results Are Better "Linux 5.4",10.313,10.154,10.298 "Linux 5.8",10.433,10.354,10.333 "Linux 5.9 Git",10.45,10.471,10.267 "ECP-CANDLE 0.3 - Benchmark: P3B1", Lower Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "ECP-CANDLE 0.3 - Benchmark: P1B2", Lower Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "ASTC Encoder 2.0 - Preset: Fast", Lower Results Are Better "Linux 5.4",9.01,9.02,9.02 "Linux 5.8",9.07,9.05,9.24 "Linux 5.9 Git",9.06,9.07,9.08 "WebP Image Encode 1.1 - Encode Settings: Quality 100", Lower Results Are Better "Linux 5.4",2.729,2.726,2.728 "Linux 5.8",2.745,2.737,2.744 "Linux 5.9 Git",2.766,2.75,2.751 "libavif avifenc 0.7.3 - Encoder Speed: 10", Lower Results Are Better "Linux 5.4",7.592,7.633,7.575 "Linux 5.8",7.612,7.606,7.7 "Linux 5.9 Git",7.662,7.634,7.738 "Mobile Neural Network 2020-09-17 - Model: inception-v3", Lower Results Are Better "Linux 5.4",62.46,62.018,62.446 "Linux 5.8",63.327,62.759,62.708 "Linux 5.9 Git",63.213,63.191,61.279 "BYTE Unix Benchmark 3.6 - Computational Test: Dhrystone 2", Higher Results Are Better "Linux 5.4",38558685.6,38621892.2,38435069.4 "Linux 5.8",38517532.9,38501253.6,38522737.6 "Linux 5.9 Git",38369371.3,37860344.2,38278301 "GPAW 20.1 - Input: Carbon Nanotube", Lower Results Are Better "Linux 5.4",599.385,598.684,596.891 "Linux 5.8",596.908,597.951,596.421 "Linux 5.9 Git",595.252,610.247,602.872 "Mlpack Benchmark - Benchmark: scikit_svm", Lower Results Are Better "Linux 5.4",28.059153795242,27.77561879158,27.750396251678 "Linux 5.8",28.378074169159,28.00993514061,27.963325738907 "Linux 5.9 Git",27.865969657898,27.872714757919,27.867217063904 "RNNoise 2020-06-28 - ", Lower Results Are Better "Linux 5.4",27.731,27.964,27.915 "Linux 5.8",27.938,28.498,27.955 "Linux 5.9 Git",28.22,28.016,28.028 "Timed Apache Compilation 2.4.41 - Time To Compile", Lower Results Are Better "Linux 5.4",32.63,32.731,32.836 "Linux 5.8",32.757,32.754,32.624 "Linux 5.9 Git",33.01,32.878,33.137 "TSCP 1.81 - AI Chess Performance", Higher Results Are Better "Linux 5.4",1140327,1140327,1142692,1137971,1145068 "Linux 5.8",1135624,1130960,1133288,1135624,1137971 "Linux 5.9 Git",1133288,1133288,1128643,1128643,1133288 "libavif avifenc 0.7.3 - Encoder Speed: 2", Lower Results Are Better "Linux 5.4",108.065,107.782,108.324 "Linux 5.8",108.504,108.505,108.201 "Linux 5.9 Git",109.32,108.786,108.761 "InfluxDB 1.8.2 - Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000", Higher Results Are Better "Linux 5.4",991986.1,988488.2,989810.9 "Linux 5.8",989213.4,985072.4,979650.1 "Linux 5.9 Git",984461.9,989205.2,972609.1 "XZ Compression 5.2.4 - Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9", Lower Results Are Better "Linux 5.4",50.83,50.693,50.682 "Linux 5.8",50.488,50.835,50.532 "Linux 5.9 Git",50.633,51.759,50.687 "WebP Image Encode 1.1 - Encode Settings: Quality 100, Lossless", Lower Results Are Better "Linux 5.4",19.386,19.296,19.402 "Linux 5.8",19.312,19.36,19.328 "Linux 5.9 Git",19.232,19.175,19.234 "NCNN 20200916 - Target: CPU - Model: mnasnet", Lower Results Are Better "Linux 5.4",6.62,6.75,6.61 "Linux 5.8",6.6,6.61,6.62 "Linux 5.9 Git",6.6,6.62,6.63 "AOM AV1 2.0 - Encoder Mode: Speed 8 Realtime", Higher Results Are Better "Linux 5.4",35.96,35.83,35.66 "Linux 5.8",35.65,35.59,35.48 "Linux 5.9 Git",35.74,35.63,35.58 "Tesseract OCR 4.1.1 - Time To OCR 7 Images", Lower Results Are Better "Linux 5.4",27.526,27.406,27.41 "Linux 5.8",27.621,27.702,27.593 "Linux 5.9 Git",27.562,27.601,27.538 "Rodinia 3.1 - Test: OpenMP LavaMD", Lower Results Are Better "Linux 5.4",576.102,567.471,567.668 "Linux 5.8",566.779,570.075,567.427 "Linux 5.9 Git",566.585,566.736,566.274 "Mobile Neural Network 2020-09-17 - Model: mobilenet-v1-1.0", Lower Results Are Better "Linux 5.4",7.763,7.733,7.747 "Linux 5.8",7.815,7.775,7.774 "Linux 5.9 Git",7.805,7.794,7.8 "NCNN 20200916 - Target: CPU - Model: efficientnet-b0", Lower Results Are Better "Linux 5.4",10.44,10.63,10.45 "Linux 5.8",10.45,10.44,10.44 "Linux 5.9 Git",10.45,10.45,10.44 "Mobile Neural Network 2020-09-17 - Model: MobileNetV2_224", Lower Results Are Better "Linux 5.4",5.612,5.564,5.628 "Linux 5.8",5.618,5.673,5.622 "Linux 5.9 Git",5.605,5.67,5.541 "Timed Linux Kernel Compilation 5.4 - Time To Compile", Lower Results Are Better "Linux 5.4",200.943,198.248,199.714 "Linux 5.8",200.961,199.173,199.233 "Linux 5.9 Git",202.512,200.308,200.023 "TNN 0.2.3 - Target: CPU - Model: SqueezeNet v1.1", Lower Results Are Better "Linux 5.4",344.311,344.602,344.419 "Linux 5.8",346.586,346.463,346.452 "Linux 5.9 Git",345.339,346.484,346.598 "NCNN 20200916 - Target: CPU - Model: vgg16", Lower Results Are Better "Linux 5.4",89.24,89.46,89.34 "Linux 5.8",88.84,89.19,88.81 "Linux 5.9 Git",88.85,88.69,88.91 "Rodinia 3.1 - Test: OpenMP CFD Solver", Lower Results Are Better "Linux 5.4",47.335,47.342,47.502 "Linux 5.8",47.56,47.72,47.733 "Linux 5.9 Git",47.629,47.619,47.646 "WebP Image Encode 1.1 - Encode Settings: Quality 100, Lossless, Highest Compression", Lower Results Are Better "Linux 5.4",47.518,47.195,47.686 "Linux 5.8",47.305,47.299,47.175 "Linux 5.9 Git",47.187,47.348,47.034 "Rodinia 3.1 - Test: OpenMP HotSpot3D", Lower Results Are Better "Linux 5.4",118.401,118.259,118.248 "Linux 5.8",118.36,119.028,117.99 "Linux 5.9 Git",119.325,118.385,119.222 "MPV - Video Input: Big Buck Bunny Sunflower 4K - Decode: Software Only", Higher Results Are Better "Linux 5.4",32.138862879345,32.114960944698,32.048252351002 "Linux 5.8",32.16818713514,32.241726661624,32.240411002574 "Linux 5.9 Git",32.255468723129,32.24165512133,32.338220197115 "Caffe 2020-02-13 - Model: GoogleNet - Acceleration: CPU - Iterations: 200", Lower Results Are Better "Linux 5.4",329252,328892,328519 "Linux 5.8",330531,330531,330242 "Linux 5.9 Git",330487,331277,330229 "AOM AV1 2.0 - Encoder Mode: Speed 4 Two-Pass", Higher Results Are Better "Linux 5.4",1.88,1.88,1.89 "Linux 5.8",1.88,1.87,1.87 "Linux 5.9 Git",1.88,1.88,1.87 "Dolfyn 0.527 - Computational Fluid Dynamics", Lower Results Are Better "Linux 5.4",21.459,21.368,21.449 "Linux 5.8",21.507,21.411,21.51 "Linux 5.9 Git",21.561,21.489,21.56 "WebP Image Encode 1.1 - Encode Settings: Default", Lower Results Are Better "Linux 5.4",1.743,1.739,1.74 "Linux 5.8",1.75,1.749,1.75 "Linux 5.9 Git",1.748,1.75,1.749 "NCNN 20200916 - Target: CPU-v2-v2 - Model: mobilenet-v2", Lower Results Are Better "Linux 5.4",7.72,7.89,7.77 "Linux 5.8",7.8,7.78,7.87 "Linux 5.9 Git",7.8,7.76,7.77 "oneDNN 1.5 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",7.15543,7.1761,7.16572 "Linux 5.8",7.17185,7.18981,7.23436 "Linux 5.9 Git",7.1642,7.22974,7.21216 "Zstd Compression 1.4.5 - Compression Level: 19", Higher Results Are Better "Linux 5.4",20.5,20.4,20.5 "Linux 5.8",20.4,20.4,20.4 "Linux 5.9 Git",20.5,20.3,20.5 "NCNN 20200916 - Target: CPU - Model: squeezenet", Lower Results Are Better "Linux 5.4",24.73,24.96,24.96 "Linux 5.8",24.72,24.73,24.83 "Linux 5.9 Git",24.74,24.81,24.97 "Incompact3D 2020-09-17 - Input: Cylinder", Lower Results Are Better "Linux 5.4",749.837158,748.988098,753.927612 "Linux 5.8",754.420715,754.080505,754.380493 "Linux 5.9 Git",753.819336,754.776367,754.212463 "WebP Image Encode 1.1 - Encode Settings: Quality 100, Highest Compression", Lower Results Are Better "Linux 5.4",8.135,8.165,8.145 "Linux 5.8",8.176,8.174,8.203 "Linux 5.9 Git",8.132,8.172,8.179 "LAMMPS Molecular Dynamics Simulator 24Aug2020 - Model: Rhodopsin Protein", Higher Results Are Better "Linux 5.4",2.755,2.744,2.74 "Linux 5.8",2.756,2.751,2.767 "Linux 5.9 Git",2.747,2.757,2.77 "libavif avifenc 0.7.3 - Encoder Speed: 0", Lower Results Are Better "Linux 5.4",183.363,182.905,181.099 "Linux 5.8",181.398,183.07,183.896 "Linux 5.9 Git",183.188,183.23,183.256 "oneDNN 1.5 - Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",18.699,18.8742,18.6856 "Linux 5.8",18.7595,18.9507,18.7754 "Linux 5.9 Git",18.751,18.896,18.7334 "LibRaw 0.20 - Post-Processing Benchmark", Higher Results Are Better "Linux 5.4",27.65,27.65,27.68 "Linux 5.8",27.55,27.72,27.69 "Linux 5.9 Git",27.59,27.49,27.58 "GROMACS 2020.3 - Water Benchmark", Higher Results Are Better "Linux 5.4",0.503,0.502,0.506 "Linux 5.8",0.502,0.504,0.505 "Linux 5.9 Git",0.508,0.502,0.495 "FFTE 7.0 - N=256, 3D Complex FFT Routine", Higher Results Are Better "Linux 5.4",20275.223279336,20196.603654225,20513.658599072 "Linux 5.8",20375.470704475,20223.679357523,20176.638597465 "Linux 5.9 Git",20431.832784028,20212.182744465,20361.912470801 "NCNN 20200916 - Target: CPU - Model: mobilenet", Lower Results Are Better "Linux 5.4",28.8,28.95,29.08 "Linux 5.8",28.93,28.79,28.84 "Linux 5.9 Git",28.75,28.77,28.99 "AOM AV1 2.0 - Encoder Mode: Speed 6 Two-Pass", Higher Results Are Better "Linux 5.4",3,3,3.01 "Linux 5.8",2.99,3,2.99 "Linux 5.9 Git",3,2.98,2.98 "Monte Carlo Simulations of Ionised Nebulae 2019-03-24 - Input: Dust 2D tau100.0", Lower Results Are Better "Linux 5.4",304,304,306 "Linux 5.8", "Linux 5.9 Git",305,303,304 "NAMD 2.14 - ATPase Simulation - 327,506 Atoms", Lower Results Are Better "Linux 5.4",3.87239,3.87397,3.87662 "Linux 5.8",3.88258,3.89567,3.88046 "Linux 5.9 Git",3.89385,3.87947,3.88777 "Caffe 2020-02-13 - Model: GoogleNet - Acceleration: CPU - Iterations: 100", Lower Results Are Better "Linux 5.4",164273,164937,164724 "Linux 5.8",165143,165118,165157 "Linux 5.9 Git",165495,165255,164794 "MPV - Video Input: Big Buck Bunny Sunflower 1080p - Decode: Software Only", Higher Results Are Better "Linux 5.4",139.50756771062,139.41902431037,139.43520147058 "Linux 5.8",139.75592549059,139.85820169579,140.02086742628 "Linux 5.9 Git",139.74374912915,139.9390200505,139.55501831815 "NCNN 20200916 - Target: CPU - Model: yolov4-tiny", Lower Results Are Better "Linux 5.4",39.38,39.6,39.47 "Linux 5.8",39.36,39.39,39.34 "Linux 5.9 Git",39.32,39.41,39.63 "InfluxDB 1.8.2 - Concurrent Streams: 1024 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000", Higher Results Are Better "Linux 5.4",996463,998256.8,994660 "Linux 5.8",992800.8,992971.9,994539.4 "Linux 5.9 Git",993468.2,999003.7,992386.9 "oneDNN 1.5 - Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",19.8836,20.0713,19.8728 "Linux 5.8",19.828,20.0314,19.7992 "Linux 5.9 Git",19.8143,20.0044,19.828 "AI Benchmark Alpha 0.1.2 - Device Inference Score", Higher Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "Caffe 2020-02-13 - Model: AlexNet - Acceleration: CPU - Iterations: 100", Lower Results Are Better "Linux 5.4",66874,67031,66921 "Linux 5.8",67090,67043,67067 "Linux 5.9 Git",66836,66980,66847 "OCRMyPDF 9.6.0+dfsg - Processing 60 Page PDF Document", Lower Results Are Better "Linux 5.4",55.804,55.362,55.447 "Linux 5.8",55.642,55.322,55.216 "Linux 5.9 Git",55.91,55.345,55.271 "NCNN 20200916 - Target: CPU - Model: resnet18", Lower Results Are Better "Linux 5.4",23.06,23.1,23.11 "Linux 5.8",23.17,23.06,23.1 "Linux 5.9 Git",23.05,23.04,23.05 "Timed LLVM Compilation 10.0 - Time To Compile", Lower Results Are Better "Linux 5.4",1507.882,1487.303,1484.481 "Linux 5.8",1492.176,1489.902,1488.902 "Linux 5.9 Git",1494.306,1493.86,1493.383 "NCNN 20200916 - Target: CPU - Model: googlenet", Lower Results Are Better "Linux 5.4",23.03,23.16,23.06 "Linux 5.8",23.03,23.02,23.03 "Linux 5.9 Git",23.03,23.05,23.13 "NCNN 20200916 - Target: CPU - Model: shufflenet-v2", Lower Results Are Better "Linux 5.4",4.62,4.71,4.67 "Linux 5.8",4.63,4.68,4.68 "Linux 5.9 Git",4.62,4.68,4.67 "Blender 2.90 - Blend File: BMW27 - Compute: CPU-Only", Lower Results Are Better "Linux 5.4",383.51,381.62,382.88 "Linux 5.8",380.82,381.49,384.43 "Linux 5.9 Git",380.87,382.47,382.46 "Timed MAFFT Alignment 7.471 - Multiple Sequence Alignment - LSU RNA", Lower Results Are Better "Linux 5.4",12.334,12.322,12.148 "Linux 5.8",12.465,12.199,12.209 "Linux 5.9 Git",12.55,12.26,12.063 "Timed HMMer Search 3.3.1 - Pfam Database Search", Lower Results Are Better "Linux 5.4",125.446,125.432,125.411 "Linux 5.8",125.536,125.372,125.577 "Linux 5.9 Git",125.791,125.593,125.597 "oneDNN 1.5 - Harness: Deconvolution Batch deconv_3d - Data Type: u8s8f32 - Engine: CPU", Lower Results Are Better "Linux 5.4",7.38779,7.45967,7.41024 "Linux 5.8",7.39113,7.45187,7.42523 "Linux 5.9 Git",7.39934,7.41749,7.4115 "Kripke 1.2.4 - ", Higher Results Are Better "Linux 5.4",17865810,17811750,17829630 "Linux 5.8",17983390,17828050,17787630 "Linux 5.9 Git",18006200,17712740,17825710 "NCNN 20200916 - Target: CPU-v3-v3 - Model: mobilenet-v3", Lower Results Are Better "Linux 5.4",6.45,6.59,6.54 "Linux 5.8",6.53,6.53,6.54 "Linux 5.9 Git",6.51,6.51,6.54 "ASTC Encoder 2.0 - Preset: Thorough", Lower Results Are Better "Linux 5.4",72.24,72.26,72.27 "Linux 5.8",72.3,72.33,72.34 "Linux 5.9 Git",72.33,72.39,72.39 "AI Benchmark Alpha 0.1.2 - Device AI Score", Higher Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "Zstd Compression 1.4.5 - Compression Level: 3", Higher Results Are Better "Linux 5.4",2373.9,2390,2394.2 "Linux 5.8",2386.1,2389,2392.9 "Linux 5.9 Git",2389.3,2386.5,2391.1 "TensorFlow Lite 2020-08-23 - Model: Mobilenet Quant", Lower Results Are Better "Linux 5.4",348856,349051,349091 "Linux 5.8",348854,349071,349002 "Linux 5.9 Git",349245,349592,349477 "Hierarchical INTegration 1.0 - Test: FLOAT", Higher Results Are Better "Linux 5.4",379948764.1013,380481390.18202,384277749.09627 "Linux 5.8",381085501.97728,380880986.44647,381845569.9488 "Linux 5.9 Git",382500270.73929,380079644.87897,380823156.9491 "Blender 2.90 - Blend File: Fishy Cat - Compute: CPU-Only", Lower Results Are Better "Linux 5.4",530.84,533.68,532.96 "Linux 5.8",530.97,532.88,531.88 "Linux 5.9 Git",533.77,531.62,531.14 "Rodinia 3.1 - Test: OpenMP Streamcluster", Lower Results Are Better "Linux 5.4",28.694,28.731,28.776 "Linux 5.8",28.664,28.742,28.785 "Linux 5.9 Git",28.633,28.717,28.758 "NCNN 20200916 - Target: CPU - Model: resnet50", Lower Results Are Better "Linux 5.4",46.55,46.45,46.5 "Linux 5.8",46.44,46.48,46.43 "Linux 5.9 Git",46.44,46.48,46.46 "TNN 0.2.3 - Target: CPU - Model: MobileNet v2", Lower Results Are Better "Linux 5.4",365.539,367.152,366.201 "Linux 5.8",366.458,366.227,366.494 "Linux 5.9 Git",366.8,366.449,366.644 "ASTC Encoder 2.0 - Preset: Medium", Lower Results Are Better "Linux 5.4",11.14,11.14,11.14 "Linux 5.8",11.14,11.14,11.14 "Linux 5.9 Git",11.14,11.16,11.15 "Caffe 2020-02-13 - Model: AlexNet - Acceleration: CPU - Iterations: 200", Lower Results Are Better "Linux 5.4",133799,133834,133836 "Linux 5.8",133792,133942,133877 "Linux 5.9 Git",133804,133926,133593 "TensorFlow Lite 2020-08-23 - Model: NASNet Mobile", Lower Results Are Better "Linux 5.4",366874,366899,367067 "Linux 5.8",366744,366883,367116 "Linux 5.9 Git",367225,367043,366965 "TensorFlow Lite 2020-08-23 - Model: SqueezeNet", Lower Results Are Better "Linux 5.4",499628,499711,499744 "Linux 5.8",499941,499755,499840 "Linux 5.9 Git",499707,499847,499922 "ASTC Encoder 2.0 - Preset: Exhaustive", Lower Results Are Better "Linux 5.4",578.23,578.11,578.18 "Linux 5.8",578.71,578.23,577.97 "Linux 5.9 Git",578.3,578.32,578.05 "TensorFlow Lite 2020-08-23 - Model: Inception ResNet V2", Lower Results Are Better "Linux 5.4",6516570,6518660,6517030 "Linux 5.8",6518270,6517970,6518170 "Linux 5.9 Git",6520390,6517430,6518560 "TensorFlow Lite 2020-08-23 - Model: Inception V4", Lower Results Are Better "Linux 5.4",7204660,7204000,7203840 "Linux 5.8",7204010,7204430,7204380 "Linux 5.9 Git",7202930,7201890,7203900 "TensorFlow Lite 2020-08-23 - Model: Mobilenet Float", Lower Results Are Better "Linux 5.4",339294,339480,339461 "Linux 5.8",339303,339375,339439 "Linux 5.9 Git",339393,339530,339379 "AI Benchmark Alpha 0.1.2 - Device Training Score", Higher Results Are Better "Linux 5.4", "Linux 5.8", "Linux 5.9 Git", "AOM AV1 2.0 - Encoder Mode: Speed 6 Realtime", Higher Results Are Better "Linux 5.4",14.96,14.96,15.04 "Linux 5.8",14.95,15.03,14.99 "Linux 5.9 Git",14.99,15.01,14.97 "AOM AV1 2.0 - Encoder Mode: Speed 0 Two-Pass", Higher Results Are Better "Linux 5.4",0.22,0.22,0.22 "Linux 5.8",0.22,0.22,0.22 "Linux 5.9 Git",0.22,0.22,0.22 "InfluxDB 1.8.2 - Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000", Higher Results Are Better "Linux 5.4",963781.8,936035.8,879630.7,861810.2,866717,861471.6,850506.6,846235.6,846263,855102.5,846953.3,837070.7 "Linux 5.8",957005.7,926529.3,851893.6,830105.3,793167.2,787822.1,776813.4,796125.9,745560.5,795638.4,752594.6,760290.7 "Linux 5.9 Git",961865.4,919584,852217.9,818053.9,793033.5,797633.2,792146.6,770117.3,749366,752528.7,748687.7,713219.8 "Mlpack Benchmark - Benchmark: scikit_qda", Lower Results Are Better "Linux 5.4",83.308103322983,104.92770528793,91.449726819992,96.515741825104,93.296894788742,86.479510068893,85.269785165787,83.315961360931,83.241715908051,83.973000526428,84.426793336868,83.979216575623 "Linux 5.8",82.713583469391,87.137753009796,84.431623220444 "Linux 5.9 Git",98.376578092575,91.195413351059,84.365674495697,83.863043069839,84.201458454132,85.534678936005,84.010852098465,86.235602140427,83.231961965561,83.692427873611,84.180478096008,84.49742436409 "NCNN 20200916 - Target: Vulkan GPU - Model: blazeface", Lower Results Are Better "Linux 5.4",2.64,2.62,2.63 "Linux 5.8",2.65,2.18,2.23 "Linux 5.9 Git",2.33,2.23,2.31 "NCNN 20200916 - Target: Vulkan GPU - Model: shufflenet-v2", Lower Results Are Better "Linux 5.4",10.13,10.09,10.12 "Linux 5.8",10.07,10.58,9.53 "Linux 5.9 Git",11.74,11.46,10.07