AMD EPYC 7003 EO 2021 Linux Comparison

2 x AMD EPYC 75F3 32-Core testing with a ASRockRack ROME2D16-2T (P3.10 BIOS) and ASPEED on AlmaLinux 8.4 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2110221-TJ-2110222TJ57&grr&sro.

GNU Radio

Test: Hilbert Transform

GNU Radio

Test: FM Deemphasis Filter

GNU Radio

Test: IIR Filter

GNU Radio

Test: FIR Filter

GNU Radio

Test: Signal Source (Cosine)

GNU Radio

Test: Five Back to Back FIR Filters

NCNN

Target: CPU - Model: regnety_400m

NCNN

Target: CPU - Model: squeezenet_ssd

NCNN

Target: CPU - Model: yolov4-tiny

NCNN

Target: CPU - Model: resnet50

NCNN

Target: CPU - Model: alexnet

NCNN

Target: CPU - Model: resnet18

NCNN

Target: CPU - Model: vgg16

NCNN

Target: CPU - Model: googlenet

NCNN

Target: CPU - Model: blazeface

NCNN

Target: CPU - Model: efficientnet-b0

NCNN

Target: CPU - Model: mnasnet

NCNN

Target: CPU - Model: shufflenet-v2

NCNN

Target: CPU-v3-v3 - Model: mobilenet-v3

NCNN

Target: CPU-v2-v2 - Model: mobilenet-v2

NCNN

Target: CPU - Model: mobilenet

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 250 - Mode: Read Write - Average Latency

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 250 - Mode: Read Write

High Performance Conjugate Gradient

Redis Memtier / Redis Benchmark

Test: SET

PlaidML

FP16: No - Mode: Inference - Network: NASNer Large - Device: CPU

ONNX Runtime

Model: shufflenet-v2-10 - Device: OpenMP CPU

WireGuard + Linux Networking Stack Stress Test

ONNX Runtime

Model: yolov4 - Device: OpenMP CPU

ONNX Runtime

Model: super-resolution-10 - Device: OpenMP CPU

Redis Memtier / Redis Benchmark

Test: GET

NWChem

Input: C240 Buckyball

Zstd Compression

Compression Level: 19, Long Mode - Decompression Speed

Zstd Compression

Compression Level: 19, Long Mode - Compression Speed

Redis Memtier / Redis Benchmark

Test: MIX

ONNX Runtime

Model: fcn-resnet101-11 - Device: OpenMP CPU

ONNX Runtime

Model: bertsquad-10 - Device: OpenMP CPU

PlaidML

FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU

Timed LLVM Compilation

Build System: Ninja

Appleseed

Scene: Material Tester

Quantum ESPRESSO

Input: AUSURF112

Timed LLVM Compilation

Build System: Unix Makefiles

TNN

Target: CPU - Model: DenseNet

Chaos Group V-RAY

Mode: CPU

OpenSSL

Algorithm: SHA256

Numpy Benchmark

PlaidML

FP16: No - Mode: Inference - Network: DenseNet 201 - Device: CPU

Zstd Compression

Compression Level: 8, Long Mode - Decompression Speed

Zstd Compression

Compression Level: 8, Long Mode - Compression Speed

PlaidML

FP16: No - Mode: Inference - Network: VGG16 - Device: CPU

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 500 - Mode: Read Write - Average Latency

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 500 - Mode: Read Write

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 500 - Mode: Read Only - Average Latency

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 500 - Mode: Read Only

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 250 - Mode: Read Only - Average Latency

PostgreSQL pgbench

Scaling Factor: 100 - Clients: 250 - Mode: Read Only

Redis Memtier / Redis Benchmark

Test: LPUSH and LPOP: lpush

VP9 libvpx Encoding

Speed: Speed 5 - Input: Bosphorus 4K

Zstd Compression

Compression Level: 8 - Decompression Speed

PyPerformance

Benchmark: raytrace

PyPerformance

Benchmark: python_startup

Blender

Blend File: Pabellon Barcelona - Compute: CPU-Only

Zstd Compression

Compression Level: 3 - Decompression Speed

Zstd Compression

Compression Level: 3, Long Mode - Decompression Speed

PlaidML

FP16: No - Mode: Inference - Network: VGG19 - Device: CPU

Zstd Compression

Compression Level: 8 - Compression Speed

Redis Memtier / Redis Benchmark

Test: LPUSH and LPOP: lpop

Zstd Compression

Compression Level: 3, Long Mode - Compression Speed

OpenSSL

Algorithm: RSA4096

OpenSSL

Algorithm: RSA4096

Timed Node.js Compilation

Time To Compile

Zstd Compression

Compression Level: 3 - Compression Speed

PyPerformance

Benchmark: go

PlaidML

FP16: No - Mode: Inference - Network: Mobilenet - Device: CPU

Zstd Compression

Compression Level: 19 - Decompression Speed

Zstd Compression

Compression Level: 19 - Compression Speed

Timed Linux Kernel Compilation

Time To Compile

GEGL

Operation: Color Enhance

ACES DGEMM

Sustained Floating-Point Rate

PyPerformance

Benchmark: crypto_pyaes

PyPerformance

Benchmark: regex_compile

PyPerformance

Benchmark: django_template

GROMACS

Implementation: MPI CPU - Input: water_GMX50_bare

Timed Wasmer Compilation

Time To Compile

GEGL

Operation: Rotate 90 Degrees

Appleseed

Scene: Disney Material

NAMD

ATPase Simulation - 327,506 Atoms

x265

Video Input: Bosphorus 4K

PyPerformance

Benchmark: float

PyPerformance

Benchmark: pathlib

GEGL

Operation: Reflect

Blender

Blend File: BMW27 - Compute: CPU-Only

PyPerformance

Benchmark: pickle_pure_python

OSPray

Demo: San Miguel - Renderer: SciVis

RawTherapee

Total Benchmark Time

PyPerformance

Benchmark: nbody

PyBench

Total For Average Test Times

PyPerformance

Benchmark: json_loads

TNN

Target: CPU - Model: MobileNet v2

OpenSSL

OpenSSL

TNN

Target: CPU - Model: SqueezeNet v1.1

Inkscape

Operation: SVG Files To PNG

Intel Open Image Denoise

Run: RT.hdr_alb_nrm.3840x2160

GEGL

Operation: Crop

AOM AV1

Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K

GEGL

Operation: Scale

OSPray

Demo: XFrog Forest - Renderer: SciVis

Nginx

Test: Long Connection - Connections: 1000

Nginx

Test: Long Connection - Connections: 500

Nginx

Test: Short Connection - Connections: 1000

Nginx

Test: Short Connection - Connections: 500

Kvazaar

Video Input: Bosphorus 4K - Video Preset: Very Fast

AOM AV1

Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K

ASTC Encoder

Preset: Exhaustive

AOM AV1

Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4K

SVT-AV1

Encoder Mode: Preset 8 - Input: Bosphorus 4K

OCRMyPDF

Processing 60 Page PDF Document

Hugin

Panorama Photo Assistant + Stitching Time

Timed Godot Game Engine Compilation

Time To Compile

SVT-VP9

Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p

POV-Ray

Trace Time

GNU Octave Benchmark