Amazon AWS Graviton3E vs. Graviton 2/3 benchmarks
Benchmarks by Michael Larabel for a future article on Phoronix.com.
HTML result view exported from: https://openbenchmarking.org/result/2308110-NE-2307106NE96&sor&grs.
NAS Parallel Benchmarks
Test / Class: LU.C
Liquid-DSP
Threads: 32 - Buffer Length: 256 - Filter Length: 512
OpenSSL
Algorithm: RSA4096
srsRAN Project
Test: Downlink Processor Benchmark
NAS Parallel Benchmarks
Test / Class: SP.C
Liquid-DSP
Threads: 64 - Buffer Length: 256 - Filter Length: 512
srsRAN Project
Test: PUSCH Processor Benchmark, Throughput Thread
OpenSSL
Algorithm: RSA4096
Liquid-DSP
Threads: 32 - Buffer Length: 256 - Filter Length: 57
Graph500
Scale: 26
Graph500
Scale: 26
OpenSSL
Algorithm: AES-256-GCM
OpenSSL
Algorithm: AES-128-GCM
ACES DGEMM
Sustained Floating-Point Rate
Stress-NG
Test: Memory Copying
Stress-NG
Test: Matrix Math
Stress-NG
Test: Vector Shuffle
nekRS
Input: Kershaw
Stress-NG
Test: Matrix 3D Math
Monte Carlo Simulations of Ionised Nebulae
Input: Dust 2D tau100.0
Xcompact3d Incompact3d
Input: input.i3d 129 Cells Per Direction
Stress-NG
Test: Vector Floating Point
OpenSSL
Algorithm: SHA512
Xcompact3d Incompact3d
Input: input.i3d 193 Cells Per Direction
Rodinia
Test: OpenMP CFD Solver
Algebraic Multi-Grid Benchmark
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512
Stress-NG
Test: Fused Multiply-Add
OpenSSL
Algorithm: ChaCha20
Graph500
Scale: 26
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 256
OpenSSL
Algorithm: ChaCha20-Poly1305
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 256
nekRS
Input: TurboPipe Periodic
NAS Parallel Benchmarks
Test / Class: MG.C
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 256
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128
LAMMPS Molecular Dynamics Simulator
Model: Rhodopsin Protein
Graph500
Scale: 26
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 128
LAMMPS Molecular Dynamics Simulator
Model: 20k Atoms
Pennant
Test: leblancbig
NWChem
Input: C240 Buckyball
Pennant
Test: sedovbig
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 256
Liquid-DSP
Threads: 64 - Buffer Length: 256 - Filter Length: 57
GROMACS
Implementation: MPI CPU - Input: water_GMX50_bare
LULESH
nginx
Connections: 500
HeFFTe - Highly Efficient FFT for Exascale
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128
NAS Parallel Benchmarks
Test / Class: CG.C
NAS Parallel Benchmarks
Test / Class: EP.D
QMCPACK
Input: simple-H2O
srsRAN Project
Test: PUSCH Processor Benchmark, Throughput Total
GPAW
Input: Carbon Nanotube
QMCPACK
Input: FeCO6_b3lyp_gms
Monte Carlo Simulations of Ionised Nebulae
Input: Gas HII40
BRL-CAD
VGR Performance Metric
LeelaChessZero
Backend: Eigen
nginx
Connections: 1000
Kripke
QMCPACK
Input: FeCO6_b3lyp_gms
Remhos
Test: Sample Remap Example
Liquid-DSP
Threads: 32 - Buffer Length: 256 - Filter Length: 32
Stress-NG
Test: Wide Vector Math
Laghos
Test: Sedov Blast Wave, ube_922_hex.mesh
Stress-NG
Test: Vector Math
Liquid-DSP
Threads: 64 - Buffer Length: 256 - Filter Length: 32
Timed Godot Game Engine Compilation
Time To Compile
LeelaChessZero
Backend: BLAS
QMCPACK
Input: Li2_STO_ae
Rodinia
Test: OpenMP LavaMD
7-Zip Compression
Test: Compression Rating
Laghos
Test: Triple Point Problem
Coremark
CoreMark Size 666 - Iterations Per Second
OpenSSL
Algorithm: SHA256
Timed Gem5 Compilation
Time To Compile
Timed Node.js Compilation
Time To Compile
7-Zip Compression
Test: Decompression Rating
Stress-NG
Test: CPU Cache
Stress-NG
Test: NUMA
Stockfish
Total Time
HeFFTe - Highly Efficient FFT for Exascale
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 128
Rodinia
Test: OpenMP Streamcluster
Phoronix Test Suite v10.8.5