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
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.
A:
Processor: 2 x AMD EPYC 75F3 32-Core @ 2.95GHz (64 Cores / 128 Threads), Motherboard: ASRockRack ROME2D16-2T (P3.30 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 1000GB Western Digital WD_BLACK SN850 1TB, Graphics: ASPEED, Audio: AMD Starship/Matisse, Network: 2 x Intel 10G X550T
OS: Ubuntu 21.10, Kernel: 5.17.0-051700rc4daily20220219-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server, Vulkan: 1.1.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1024x768
AA:
Processor: 2 x AMD EPYC 75F3 32-Core @ 2.95GHz (64 Cores / 128 Threads), Motherboard: ASRockRack ROME2D16-2T (P3.30 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 1000GB Western Digital WD_BLACK SN850 1TB, Graphics: ASPEED, Audio: AMD Starship/Matisse, Network: 2 x Intel 10G X550T
OS: Ubuntu 21.10, Kernel: 5.17.0-051700rc4daily20220219-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server, Vulkan: 1.1.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1024x768
B:
Processor: 2 x AMD EPYC 75F3 32-Core @ 2.95GHz (64 Cores / 128 Threads), Motherboard: ASRockRack ROME2D16-2T (P3.30 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 1000GB Western Digital WD_BLACK SN850 1TB, Graphics: ASPEED, Audio: AMD Starship/Matisse, Network: 2 x Intel 10G X550T
OS: Ubuntu 21.10, Kernel: 5.17.0-051700rc4daily20220219-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server, Vulkan: 1.1.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1024x768
C:
Processor: 2 x AMD EPYC 75F3 32-Core @ 2.95GHz (64 Cores / 128 Threads), Motherboard: ASRockRack ROME2D16-2T (P3.30 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 1000GB Western Digital WD_BLACK SN850 1TB, Graphics: ASPEED, Audio: AMD Starship/Matisse, Network: 2 x Intel 10G X550T
OS: Ubuntu 21.10, Kernel: 5.17.0-051700rc4daily20220219-generic (x86_64), Desktop: GNOME Shell 40.5, Display Server: X Server, Vulkan: 1.1.182, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1024x768
oneDNN 2.6
Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 3874.18 |========================================================
AA . 4530.24 |=================================================================
B .. 4336.25 |==============================================================
C .. 4224.87 |=============================================================
oneDNN 2.6
Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 3819.45 |======================================================
AA . 4416.22 |==============================================================
B .. 4623.03 |=================================================================
C .. 4390.23 |==============================================================
oneDNN 2.6
Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 1624.25 |====================================================
AA . 2016.94 |=================================================================
B .. 1801.29 |==========================================================
C .. 1696.60 |=======================================================
oneDNN 2.6
Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 1693.03 |=============================================================
AA . 1803.51 |=================================================================
B .. 1810.37 |=================================================================
C .. 1713.39 |==============================================================
ONNX Runtime 1.11
Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 123 |====================================================================
B .. 123 |====================================================================
C .. 124 |=====================================================================
ONNX Runtime 1.11
Model: bertsquad-12 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 594 |=====================================================================
B .. 592 |=====================================================================
C .. 592 |=====================================================================
ONNX Runtime 1.11
Model: GPT-2 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 13234 |===================================================================
B .. 13106 |==================================================================
C .. 9094 |==============================================
ONNX Runtime 1.11
Model: yolov4 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 434 |=====================================================================
B .. 436 |=====================================================================
C .. 436 |=====================================================================
ONNX Runtime 1.11
Model: bertsquad-12 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 540 |==========================================================
B .. 643 |=====================================================================
C .. 539 |==========================================================
ONNX Runtime 1.11
Model: GPT-2 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 1308 |====================================================================
B .. 1225 |================================================================
C .. 1242 |=================================================================
ONNX Runtime 1.11
Model: fcn-resnet101-11 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 170 |=====================================================================
B .. 158 |================================================================
C .. 152 |==============================================================
ONNX Runtime 1.11
Model: yolov4 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 355 |=====================================================================
B .. 291 |=========================================================
C .. 316 |=============================================================
ONNX Runtime 1.11
Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 1225 |===================================================================
B .. 1235 |====================================================================
C .. 1229 |====================================================================
ONNX Runtime 1.11
Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 1194 |===============================================================
B .. 1244 |==================================================================
C .. 1279 |====================================================================
ONNX Runtime 1.11
Model: super-resolution-10 - Device: CPU - Executor: Parallel
Inferences Per Minute > Higher Is Better
AA . 4320 |====================================================================
B .. 4242 |===================================================================
C .. 4233 |===================================================================
ONNX Runtime 1.11
Model: super-resolution-10 - Device: CPU - Executor: Standard
Inferences Per Minute > Higher Is Better
AA . 4626 |=========================================
B .. 7762 |====================================================================
C .. 4488 |=======================================
perf-bench
Benchmark: Epoll Wait
ops/sec > Higher Is Better
A .. 2352 |==============================================================
AA . 2299 |=============================================================
B .. 2572 |====================================================================
C .. 2062 |=======================================================
libavif avifenc 0.10
Encoder Speed: 0
Seconds < Lower Is Better
A .. 69.46 |==================================================================
AA . 69.41 |==================================================================
B .. 70.33 |===================================================================
C .. 70.65 |===================================================================
oneDNN 2.6
Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 1.68391 |===========================================================
AA . 1.07515 |======================================
B .. 1.85300 |=================================================================
C .. 1.23753 |===========================================
oneDNN 2.6
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
AA . 4276.70 |==============================================================
B .. 4464.12 |=================================================================
C .. 3862.48 |========================================================
oneDNN 2.6
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
AA . 1711.61 |================================================================
B .. 1734.92 |=================================================================
C .. 1721.12 |================================================================
oneDNN 2.6
Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 2.94158 |==================================================
AA . 3.77920 |================================================================
B .. 3.18091 |======================================================
C .. 3.85444 |=================================================================
oneDNN 2.6
Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 4.05569 |=======================================================
AA . 4.25865 |==========================================================
B .. 4.77292 |=================================================================
C .. 2.55861 |===================================
libavif avifenc 0.10
Encoder Speed: 2
Seconds < Lower Is Better
A .. 38.78 |==================================================================
AA . 38.73 |==================================================================
B .. 37.57 |================================================================
C .. 39.57 |===================================================================
oneDNN 2.6
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 11.51 |====================================
AA . 17.55 |======================================================
B .. 12.50 |=======================================
C .. 21.67 |===================================================================
perf-bench
Benchmark: Futex Lock-Pi
ops/sec > Higher Is Better
A .. 77 |===================================================================
AA . 77 |===================================================================
B .. 80 |======================================================================
C .. 77 |===================================================================
perf-bench
Benchmark: Futex Hash
ops/sec > Higher Is Better
A .. 2976792 |=================================================================
AA . 2976712 |=================================================================
B .. 2979134 |=================================================================
C .. 2945494 |================================================================
oneDNN 2.6
Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 0.747084 |===========================================================
AA . 0.815632 |================================================================
B .. 0.757826 |===========================================================
C .. 0.743010 |==========================================================
oneDNN 2.6
Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 11.24 |===================================================================
AA . 11.14 |==================================================================
B .. 11.22 |===================================================================
C .. 11.19 |===================================================================
oneDNN 2.6
Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 0.897288 |==============================================
AA . 1.247010 |================================================================
B .. 0.928893 |================================================
C .. 0.784430 |========================================
perf-bench
Benchmark: Sched Pipe
ops/sec > Higher Is Better
A .. 345990 |=================================================================
AA . 351368 |==================================================================
B .. 346133 |=================================================================
C .. 339461 |================================================================
libavif avifenc 0.10
Encoder Speed: 6, Lossless
Seconds < Lower Is Better
A .. 7.215 |================================================================
AA . 7.389 |==================================================================
B .. 7.514 |===================================================================
C .. 7.185 |================================================================
perf-bench
Benchmark: Memcpy 1MB
GB/sec > Higher Is Better
A .. 42.87 |=================================================================
AA . 42.83 |=================================================================
B .. 44.15 |===================================================================
C .. 42.93 |=================================================================
oneDNN 2.6
Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 2.63912 |=================================================================
AA . 2.54811 |===============================================================
B .. 2.56643 |===============================================================
C .. 2.54404 |===============================================================
perf-bench
Benchmark: Memset 1MB
GB/sec > Higher Is Better
A .. 62.46 |==================================================================
AA . 63.42 |===================================================================
B .. 63.79 |===================================================================
C .. 63.19 |==================================================================
oneDNN 2.6
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
AA . 22.91 |=====================================================
B .. 27.34 |===============================================================
C .. 29.10 |===================================================================
oneDNN 2.6
Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU
ms < Lower Is Better
A .. 0.878553 |==============================================================
AA . 0.904662 |================================================================
B .. 0.803426 |=========================================================
C .. 0.829217 |===========================================================
oneDNN 2.6
Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 0.705380 |================================================================
AA . 0.708812 |================================================================
B .. 0.684799 |==============================================================
C .. 0.689583 |==============================================================
perf-bench
Benchmark: Syscall Basic
ops/sec > Higher Is Better
A .. 17051621 |================================================================
AA . 17052202 |================================================================
B .. 16941879 |================================================================
C .. 16924594 |================================================================
libavif avifenc 0.10
Encoder Speed: 10, Lossless
Seconds < Lower Is Better
A .. 5.010 |================================================================
AA . 5.229 |===================================================================
B .. 5.197 |===================================================================
C .. 4.971 |================================================================
libavif avifenc 0.10
Encoder Speed: 6
Seconds < Lower Is Better
A .. 4.278 |===================================================================
AA . 4.093 |================================================================
B .. 3.986 |==============================================================
C .. 4.215 |==================================================================
oneDNN 2.6
Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU
ms < Lower Is Better
A .. 2.12767 |=================================================================
AA . 2.10407 |================================================================
B .. 1.94632 |===========================================================
C .. 1.99639 |=============================================================
oneDNN 2.6
Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
oneDNN 2.6
Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
oneDNN 2.6
Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
oneDNN 2.6
Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
oneDNN 2.6
Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
oneDNN 2.6
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better