Ubuntu 21.10 EPYC Milan

AMD EPYC 74F3 24-Core testing with a ASRockRack ROME2D16-2T (P3.10 BIOS) and ASPEED on Ubuntu 21.04 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2110131-TJ-2110124TJ87&grs&rdt.

Stress-NG

Test: IO_uring

Cpuminer-Opt

Algorithm: Garlicoin

Stress-NG

Test: System V Message Passing

Stress-NG

Test: NUMA

Stress-NG

Test: Socket Activity

PostgreSQL pgbench

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

PostgreSQL pgbench

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

Stress-NG

Test: Forking

PostgreSQL pgbench

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

PostgreSQL pgbench

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

Cpuminer-Opt

Algorithm: Myriad-Groestl

Stress-NG

Test: Context Switching

libavif avifenc

Encoder Speed: 10

nginx

Concurrent Requests: 20

Stress-NG

Test: MMAP

Apache HTTP Server

Concurrent Requests: 20

nginx

Concurrent Requests: 100

nginx

Concurrent Requests: 200

nginx

Concurrent Requests: 500

NAS Parallel Benchmarks

Test / Class: IS.D

nginx

Concurrent Requests: 1000

Stress-NG

Test: Vector Math

Rodinia

Test: OpenMP Leukocyte

Mobile Neural Network

Model: MobileNetV2_224

Google SynthMark

Test: VoiceMark_100

Mobile Neural Network

Model: inception-v3

Mobile Neural Network

Model: SqueezeNetV1.0

libavif avifenc

Encoder Speed: 10, Lossless

PlaidML

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

Apache HTTP Server

Concurrent Requests: 100

LeelaChessZero

Backend: BLAS

MariaDB

Clients: 512

Redis

Test: GET

Apache HTTP Server

Concurrent Requests: 1000

VP9 libvpx Encoding

Speed: Speed 5 - Input: Bosphorus 4K

x265

Video Input: Bosphorus 4K

Cpuminer-Opt

Algorithm: Blake-2 S

Stress-NG

Test: MEMFD

Stress-NG

Test: Glibc Qsort Data Sorting

PlaidML

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

NAS Parallel Benchmarks

Test / Class: CG.C

VP9 libvpx Encoding

Speed: Speed 0 - Input: Bosphorus 4K

Mobile Neural Network

Model: squeezenetv1.1

Rodinia

Test: OpenMP LavaMD

JPEG XL libjxl

Input: JPEG - Encode Speed: 7

Stress-NG

Test: Malloc

TNN

Target: CPU - Model: MobileNet v2

libavif avifenc

Encoder Speed: 6

NCNN

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

PostgreSQL pgbench

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

PostgreSQL pgbench

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

NCNN

Target: CPU - Model: regnety_400m

NCNN

Target: CPU - Model: resnet50

NCNN

Target: CPU - Model: vgg16

JPEG XL libjxl

Input: JPEG - Encode Speed: 5

NCNN

Target: CPU - Model: resnet18

Apache HTTP Server

Concurrent Requests: 500

Mobile Neural Network

Model: mobilenet-v1-1.0

NCNN

Target: CPU - Model: mobilenet

Apache HTTP Server

Concurrent Requests: 1

x265

Video Input: Bosphorus 1080p

Apache HTTP Server

Concurrent Requests: 200

Xcompact3d Incompact3d

Input: input.i3d 129 Cells Per Direction

Stress-NG

Test: SENDFILE

nginx

Concurrent Requests: 1

Stress-NG

Test: CPU Stress

NCNN

Target: CPU - Model: yolov4-tiny

simdjson

Throughput Test: PartialTweets

simdjson

Throughput Test: LargeRandom

NAS Parallel Benchmarks

Test / Class: MG.C

LAMMPS Molecular Dynamics Simulator

Model: Rhodopsin Protein

TensorFlow Lite

Model: NASNet Mobile

ONNX Runtime

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

NCNN

Target: CPU - Model: googlenet

NAS Parallel Benchmarks

Test / Class: LU.C

JPEG XL libjxl

Input: PNG - Encode Speed: 8

OSPray

Demo: NASA Streamlines - Renderer: SciVis

JPEG XL libjxl

Input: JPEG - Encode Speed: 8

TNN

Target: CPU - Model: SqueezeNet v2

NCNN

Target: CPU - Model: squeezenet_ssd

JPEG XL libjxl

Input: PNG - Encode Speed: 7

JPEG XL libjxl

Input: PNG - Encode Speed: 5

OpenFOAM

Input: Motorbike 30M

NCNN

Target: CPU - Model: efficientnet-b0

Stress-NG

Test: CPU Cache

NCNN

Target: CPU - Model: mnasnet

Stress-NG

Test: Glibc C String Functions

TNN

Target: CPU - Model: DenseNet

simdjson

Throughput Test: DistinctUserID

Timed Godot Game Engine Compilation

Time To Compile

TTSIOD 3D Renderer

Phong Rendering With Soft-Shadow Mapping

AOM AV1

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

Timed Linux Kernel Compilation

Time To Compile

NAS Parallel Benchmarks

Test / Class: SP.C

NAS Parallel Benchmarks

Test / Class: SP.B

ONNX Runtime

Model: yolov4 - Device: OpenMP CPU

Rodinia

Test: OpenMP CFD Solver

ASTC Encoder

Preset: Thorough

Timed Node.js Compilation

Time To Compile

libavif avifenc

Encoder Speed: 0

Stress-NG

Test: Matrix Math

libavif avifenc

Encoder Speed: 2

Cpuminer-Opt

Algorithm: LBC, LBRY Credits

Stress-NG

Test: Memory Copying

ONNX Runtime

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

libavif avifenc

Encoder Speed: 6, Lossless

7-Zip Compression

Compress Speed Test

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Only

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average Latency

LuxCoreRender

Scene: LuxCore Benchmark - Acceleration: CPU

Facebook RocksDB

Test: Read Random Write Random

SVT-VP9

Tuning: VMAF Optimized - Input: Bosphorus 1080p

PostgreSQL pgbench

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

simdjson

Throughput Test: Kostya

LuxCoreRender

Scene: Danish Mood - Acceleration: CPU

SecureMark

Benchmark: SecureMark-TLS

NCNN

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

PostgreSQL pgbench

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

TensorFlow Lite

Model: Mobilenet Float

LAMMPS Molecular Dynamics Simulator

Model: 20k Atoms

SVT-AV1

Encoder Mode: Preset 8 - Input: Bosphorus 4K

SVT-HEVC

Tuning: 10 - Input: Bosphorus 1080p

Cpuminer-Opt

Algorithm: x25x

Chaos Group V-RAY

Mode: CPU

PlaidML

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

Facebook RocksDB

Test: Update Random

NAS Parallel Benchmarks

Test / Class: FT.C

LeelaChessZero

Backend: Eigen

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU

TensorFlow Lite

Model: Mobilenet Quant