AMD AOCC 4.0 Benchmarks

AMD Ryzen 9 7950X compiler benchmarks by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2211152-PTS-AMDAOCC460&sor&grs.

Caffe

Model: AlexNet - Acceleration: CPU - Iterations: 100

TNN

Target: CPU - Model: DenseNet

TNN

Target: CPU - Model: SqueezeNet v1.1

TNN

Target: CPU - Model: MobileNet v2

Caffe

Model: GoogleNet - Acceleration: CPU - Iterations: 100

NCNN

Target: CPU - Model: regnety_400m

TNN

Target: CPU - Model: SqueezeNet v2

ASTC Encoder

Preset: Fast

C-Ray

Total Time - 4K, 16 Rays Per Pixel

eSpeak-NG Speech Engine

Text-To-Speech Synthesis

WebP Image Encode

Encode Settings: Quality 100, Highest Compression

JPEG XL libjxl

Input: JPEG - Quality: 90

JPEG XL libjxl

Input: PNG - Quality: 90

GraphicsMagick

Operation: Sharpen

Liquid-DSP

Threads: 16 - Buffer Length: 256 - Filter Length: 57

NCNN

Target: CPU - Model: blazeface

simdjson

Throughput Test: DistinctUserID

NCNN

Target: CPU - Model: shufflenet-v2

JPEG XL Decoding libjxl

CPU Threads: 1

Crypto++

Test: Unkeyed Algorithms

Crypto++

Test: Keyed Algorithms

simdjson

Throughput Test: PartialTweets

Kripke

NCNN

Target: CPU - Model: mnasnet

ASTC Encoder

Preset: Medium

NCNN

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

NCNN

Target: CPU - Model: efficientnet-b0

GraphicsMagick

Operation: HWB Color Space

libavif avifenc

Encoder Speed: 6, Lossless

TSCP

AI Chess Performance

Coremark

CoreMark Size 666 - Iterations Per Second

OpenJPEG

Encode: NASA Curiosity Panorama M34

NCNN

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

simdjson

Throughput Test: LargeRandom

AOM AV1

Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K

JPEG XL Decoding libjxl

CPU Threads: All

FLAC Audio Encoding

WAV To FLAC

Liquid-DSP

Threads: 32 - Buffer Length: 256 - Filter Length: 57

CppPerformanceBenchmarks

Test: Stepanov Vector

GraphicsMagick

Operation: Noise-Gaussian

Xsbench

Kvazaar

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

libavif avifenc

Encoder Speed: 10, Lossless

GraphicsMagick

Operation: Rotate

LeelaChessZero

Backend: Eigen

POV-Ray

Trace Time

Google Draco

Model: Lion

libavif avifenc

Encoder Speed: 2

WebP Image Encode

Encode Settings: Quality 100, Lossless

Kvazaar

Video Input: Bosphorus 1080p - Video Preset: Very Fast

CLOMP

Static OMP Speedup

Google Draco

Model: Church Facade

Ngspice

Circuit: C2670

NCNN

Target: CPU - Model: googlenet

NCNN

Target: CPU - Model: squeezenet_ssd

SecureMark

Benchmark: SecureMark-TLS

AOBench

Size: 2048 x 2048 - Total Time

GraphicsMagick

Operation: Enhanced

OpenSSL

Algorithm: SHA256

simdjson

Throughput Test: TopTweet

CppPerformanceBenchmarks

Test: Stepanov Abstraction

yquake2

Renderer: Software CPU Color Light - AF: On - MSAA: Off - Resolution: 1920 x 1080

libavif avifenc

Encoder Speed: 6

yquake2

Renderer: Software CPU Color Light - AF: Off - MSAA: Off - Resolution: 1920 x 1080

libavif avifenc

Encoder Speed: 0

WebP Image Encode

Encode Settings: Quality 100, Lossless, Highest Compression

NCNN

Target: CPU - Model: alexnet

QuadRay

Scene: 5 - Resolution: 4K

yquake2

Renderer: Software CPU Color Light - AF: On - MSAA: On - Resolution: 1920 x 1080

QuadRay

Scene: 5 - Resolution: 1080p

LAME MP3 Encoding

WAV To MP3

NCNN

Target: CPU - Model: mobilenet

yquake2

Renderer: Software CPU Color Light - AF: Off - MSAA: On - Resolution: 1920 x 1080

JPEG XL libjxl

Input: PNG - Quality: 100

CppPerformanceBenchmarks

Test: Ctype

C-Blosc

Test: blosclz bitshuffle

simdjson

Throughput Test: Kostya

oneDNN

Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU

Ngspice

Circuit: C7552

Kvazaar

Video Input: Bosphorus 1080p - Video Preset: Ultra Fast

yquake2

Renderer: Software CPU - AF: Off - MSAA: On - Resolution: 1920 x 1080

yquake2

Renderer: Software CPU - AF: Off - MSAA: Off - Resolution: 1920 x 1080

GraphicsMagick

Operation: Swirl

yquake2

Renderer: Software CPU - AF: On - MSAA: Off - Resolution: 1920 x 1080

NCNN

Target: CPU - Model: yolov4-tiny

NCNN

Target: CPU - Model: vgg16

WebP Image Encode

Encode Settings: Quality 100

Kvazaar

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

yquake2

Renderer: Software CPU - AF: On - MSAA: On - Resolution: 1920 x 1080

NCNN

Target: CPU - Model: resnet50

AOM AV1

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

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU

WebP Image Encode

Encode Settings: Default

C-Blosc

Test: blosclz shuffle

SQLite Speedtest

Timed Time - Size 1,000

oneDNN

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU

LAMMPS Molecular Dynamics Simulator

Model: Rhodopsin Protein

GraphicsMagick

Operation: Resizing

ONNX Runtime

Model: super-resolution-10 - Device: CPU - Executor: Parallel

Redis

Test: SET - Parallel Connections: 50

NCNN

Target: CPU - Model: resnet18

Zstd Compression

Compression Level: 19 - Compression Speed

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU

Zstd Compression

Compression Level: 19, Long Mode - Decompression Speed

LeelaChessZero

Backend: BLAS

CppPerformanceBenchmarks

Test: Function Objects

ASTC Encoder

Preset: Thorough

Monte Carlo Simulations of Ionised Nebulae

Input: Dust 2D tau100.0

CppPerformanceBenchmarks

Test: Math Library

KTX-Software toktx

Settings: Zstd Compression 19