2 x AMD EPYC 75F3 32-Core testing with a ASRockRack ROME2D16-2T (P3.30 BIOS) and ASPEED on Ubuntu 21.10 via the Phoronix Test Suite.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2204097-NE-EPYC75F3N46
epyc-75f3-new ,
"ONNX Runtime 1.11 - Model: super-resolution-10 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",
"B",7761.5
"C",4487.5
"ONNX Runtime 1.11 - Model: GPT-2 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",13233.5
"B",
"C",
"oneDNN 2.6 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"AA",
"B",
"C",
"perf-bench - Benchmark: Epoll Wait",
Higher Results Are Better
"A",2590,2285,2306,2323,2354,2326,2366,2285,2371,2317,2305,2392
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: yolov4 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",354.5
"B",290.5
"C",315.5
"ONNX Runtime 1.11 - Model: bertsquad-12 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",
"B",
"C",538.5
"oneDNN 2.6 - Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",0.859594,0.829311,0.812538,0.893033,0.903891,0.849211,0.885907,0.893027,0.917892,0.933252,0.924975,0.839999
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: fcn-resnet101-11 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",169.5
"B",
"C",151.5
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",2.14122,2.07908,2.16272
"AA",
"B",
"C",
"libavif avifenc 0.10 - Encoder Speed: 6",
Lower Results Are Better
"A",4.281,4.258,4.296
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard",
Higher Results Are Better
"AA",
"B",1243.5
"C",1278.5
"ONNX Runtime 1.11 - Model: GPT-2 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",1307.5
"B",
"C",1241.5
"libavif avifenc 0.10 - Encoder Speed: 2",
Lower Results Are Better
"A",38.61,38.768,38.975
"AA",
"B",
"C",
"libavif avifenc 0.10 - Encoder Speed: 10, Lossless",
Lower Results Are Better
"A",4.966,4.997,5.066
"AA",
"B",
"C",
"libavif avifenc 0.10 - Encoder Speed: 6, Lossless",
Lower Results Are Better
"A",7.542,7.029,7.245,7.23,7.028,7.168,7.262
"AA",
"B",
"C",
"perf-bench - Benchmark: Futex Lock-Pi",
Higher Results Are Better
"A",76,77,78
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",2.60899,2.64608,2.66229
"AA",
"B",
"C",
"perf-bench - Benchmark: Sched Pipe",
Higher Results Are Better
"A",349621,339445,348905
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",0.718883,0.7013,0.695957
"AA",
"B",
"C",
"perf-bench - Benchmark: Memcpy 1MB",
Higher Results Are Better
"A",42.596511,43.171952,42.836705
"AA",
"B",
"C",
"perf-bench - Benchmark: Memset 1MB",
Higher Results Are Better
"A",62.17391,63.236293,60.027741,63.422951,63.414467
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: super-resolution-10 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",4319.5
"B",
"C",
"libavif avifenc 0.10 - Encoder Speed: 0",
Lower Results Are Better
"A",69.426,70.078,68.877
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"AA",
"B",
"C",
"perf-bench - Benchmark: Futex Hash",
Higher Results Are Better
"A",2982921,2983893,2963563
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",11.2839,11.0667,11.3577
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",1224.5
"B",1234.5
"C",1228.5
"ONNX Runtime 1.11 - Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",
"B",
"C",123.5
"perf-bench - Benchmark: Syscall Basic",
Higher Results Are Better
"A",17066614,17045104,17043146
"AA",
"B",
"C",
"ONNX Runtime 1.11 - Model: yolov4 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",433.5
"B",
"C",
"ONNX Runtime 1.11 - Model: bertsquad-12 - Device: CPU - Executor: Parallel",
Higher Results Are Better
"AA",
"B",
"C",591.5
"oneDNN 2.6 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",6.79886,8.84473,17.3921,8.9048,14.2526,9.16272,6.06559,16.9887,7.44543,14.1672,6.53782,21.2162,8.50531,9.34433,17.0557
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",1896.57,1813.43,1932.32,1799.56,1595.07,1705.25,1201.79,974.637,1566.64,1379.95,1964.09,1661.64
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"A",
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"A",
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"A",
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",4388.73,3517.76,4498,3658.74,4058.55,4426.84,3364,4510.71,4519.05,3329.94,1582.95,3978.08
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",1612.97,1704.33,1757.93,1475.88,1705.78,1777.9,1751.79,1903.75,2021.02,1582.24,1869.41,1153.4
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",4166.03,1938.1,2038,3758.08,4437.52,4494.99,4187.84,4375.8,4393.51,4220.07,2942.83,4347.81,4513.42,4424.53
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",1.9979,1.57358,2.07578,1.99527,1.97459,1.67451,1.63069,1.30021,1.54157,1.86138,1.80011,1.01819,1.35517,1.18376,2.2759
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",0.904756,0.721012,1.2437,0.651952,0.930928,0.884634,0.809588,1.15296,0.905567,0.845005,0.605328,1.11305,0.850607,1.10735,0.732877
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"A",
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU",
Lower Results Are Better
"A",
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",0.682458,0.68876,0.736981,0.775459,0.777069,0.794007,0.775001,0.74676,0.723708,0.719387,0.688051,0.697038,0.795927,0.804712,0.800941
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU",
Lower Results Are Better
"A",5.06894,4.24678,4.20127,4.04927,4.59104,3.76035,4.44797,3.4684,3.22395,4.19395,4.45327,3.34854,3.24829,3.87961,4.6537
"AA",
"B",
"C",
"oneDNN 2.6 - Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU",
Lower Results Are Better
"A",3.16136,2.37051,3.99677,3.82056,2.11965,2.63391,2.60135,2.62118,2.21314,4.16647,2.40874,2.92918,1.87171,3.0823,4.1268
"AA",
"B",
"C",