AMD Ryzen 9 7900 DDR5-5200 ECC memory benchmarks by Michael Larabel for a future article
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2405045-NE-AMDRYZEN948 AMD Ryzen 9 7900 DDR5-5200 ECC Memory - Phoronix Test Suite AMD Ryzen 9 7900 DDR5-5200 ECC Memory AMD Ryzen 9 7900 DDR5-5200 ECC memory benchmarks by Michael Larabel for a future article
HTML result view exported from: https://openbenchmarking.org/result/2405045-NE-AMDRYZEN948&sgm&sor&grr .
AMD Ryzen 9 7900 DDR5-5200 ECC Memory Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server Compiler File-System Screen Resolution 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 AMD Ryzen 9 7900 12-Core @ 5.48GHz (12 Cores / 24 Threads) ASRockRack 1U4LW-B650/2L2T B650D4U-2L2T/BCM (2.09 BIOS) AMD Device 14d8 2 x 32GB DDR5-4800MT/s Micron MTC20C2085S1EC48BA1 1024GB SOLIDIGM SSDPFKNU010TZ ASPEED AMD Rembrandt Radeon HD Audio VA2431 2 x Intel I210 + 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA Ubuntu 24.04 6.8.0-31-generic (x86_64) GNOME Shell 46.0 X Server GCC 13.2.0 ext4 1920x1200 2 x 32GB DDR5-5200MT/s Kingston OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Processor Details - Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601203 Python Details - Python 3.12.3 Security Details - gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Vulnerable: Safe RET no microcode + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
AMD Ryzen 9 7900 DDR5-5200 ECC Memory build-linux-kernel: allmodconfig openradioss: Chrysler Neon 1M brl-cad: VGR Performance Metric blender: Barbershop - CPU-Only build-nodejs: Time To Compile llama-cpp: llama-2-70b-chat.Q5_0.gguf openradioss: INIVOL and Fluid Structure Interaction Drop Container build-godot: Time To Compile blender: Pabellon Barcelona - CPU-Only openfoam: drivaerFastback, Small Mesh Size - Execution Time openfoam: drivaerFastback, Small Mesh Size - Mesh Time quicksilver: CORAL2 P2 llamafile: wizardcoder-python-34b-v1.0.Q6_K - CPU tensorflow: CPU - 64 - ResNet-50 blender: Classroom - CPU-Only openradioss: Bird Strike on Windshield pgbench: 100 - 1000 - Read Only - Average Latency pgbench: 100 - 1000 - Read Only pgbench: 100 - 1000 - Read Write - Average Latency pgbench: 100 - 1000 - Read Write npb: SP.C ffmpeg: libx265 - Platform ffmpeg: libx265 - Video On Demand ffmpeg: libx265 - Upload llamafile: mistral-7b-instruct-v0.2.Q8_0 - CPU openradioss: Bumper Beam specfem3d: Layered Halfspace specfem3d: Water-layered Halfspace blender: Junkshop - CPU-Only stockfish: Chess Benchmark blender: Fishy Cat - CPU-Only namd: STMV with 1,066,628 Atoms nginx: 1000 openradioss: Rubber O-Ring Seal Installation gromacs: MPI CPU - water_GMX50_bare incompact3d: input.i3d 193 Cells Per Direction npb: BT.C build-linux-kernel: defconfig v-ray: CPU openradioss: Cell Phone Drop Test llama-cpp: llama-2-13b.Q4_0.gguf ospray: gravity_spheres_volume/dim_512/ao/real_time tensorflow: CPU - 64 - GoogLeNet openvino: Face Detection FP16-INT8 - CPU openvino: Face Detection FP16-INT8 - CPU openvino: Machine Translation EN To DE FP16 - CPU openvino: Machine Translation EN To DE FP16 - CPU openvino: Person Detection FP16 - CPU openvino: Person Detection FP16 - CPU openvino: Person Detection FP32 - CPU openvino: Person Detection FP32 - CPU openvino: Road Segmentation ADAS FP16-INT8 - CPU openvino: Road Segmentation ADAS FP16-INT8 - CPU openvino: Person Vehicle Bike Detection FP16 - CPU openvino: Person Vehicle Bike Detection FP16 - CPU quicksilver: CORAL2 P1 openvino: Road Segmentation ADAS FP16 - CPU openvino: Road Segmentation ADAS FP16 - CPU openvino: Vehicle Detection FP16 - CPU openvino: Vehicle Detection FP16 - CPU graphics-magick: Noise-Gaussian graphics-magick: Rotate graphics-magick: Swirl graphics-magick: HWB Color Space graphics-magick: Resizing rocksdb: Read While Writing mbw: Memory Copy, Fixed Block Size - 8192 MiB libxsmm: 32 mbw: Memory Copy - 8192 MiB npb: LU.C specfem3d: Homogeneous Halfspace build-php: Time To Compile avifenc: 2 ffmpeg: libx265 - Live embree: Pathtracer ISPC - Asian Dragon Obj namd: ATPase with 327,506 Atoms specfem3d: Mount St. Helens specfem3d: Tomographic Model npb: IS.D jpegxl-decode: All ramspeed: Average - Integer ramspeed: Copy - Integer ramspeed: Scale - Integer ramspeed: Triad - Integer ramspeed: Add - Integer jpegxl-decode: 1 embree: Pathtracer ISPC - Asian Dragon jpegxl: PNG - 90 tensorflow: CPU - 64 - AlexNet mbw: Memory Copy, Fixed Block Size - 4096 MiB mbw: Memory Copy - 4096 MiB llamafile: llava-v1.5-7b-q4 - CPU pytorch: CPU - 1 - ResNet-50 llama-cpp: llama-2-7b.Q4_0.gguf build-mesa: Time To Compile compress-7zip: Decompression Rating compress-7zip: Compression Rating npb: SP.B npb: FT.C incompact3d: input.i3d 129 Cells Per Direction npb: CG.C srsran: PUSCH Processor Benchmark, Throughput Total tensorflow: CPU - 1 - ResNet-50 tensorflow: CPU - 1 - AlexNet onednn: IP Shapes 3D - CPU npb: MG.C rnnoise: 26 Minute Long Talking Sample mbw: Memory Copy, Fixed Block Size - 1024 MiB mbw: Memory Copy - 1024 MiB draco: Church Facade onednn: Convolution Batch Shapes Auto - CPU srsran: PDSCH Processor Benchmark, Throughput Total draco: Lion svt-av1: Preset 12 - Bosphorus 4K svt-av1: Preset 13 - Bosphorus 4K avifenc: 6 encode-wavpack: WAV To WavPack tensorflow: CPU - 1 - GoogLeNet avifenc: 10, Lossless onednn: Deconvolution Batch shapes_3d - CPU 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 979.122 913.48 280941 676.26 515.333 1.22 338.34 240.329 233.85 196.382 28.682586 22580000 2.02 35.44 193.27 182.19 1.385 721872 29.47 33933 11331.51 68.30 68.27 33.66 7.21 105.47 102.828797179 100.601281906 99.78 30072838 94.72 0.62007 114755.2 83.39 2.151 79.4299469 36869.25 76.578 31844 65.18 7.3 5.49836 112.13 322.63 18.57 60.53 99.05 80.58 74.4 80.04 74.91 17.32 345.8 5.48 1090.85 18550000 21.27 281.64 9.26 645.63 91 165 187 259 289 3174105 16405.968 107.1 17709.526 41111.39 47.018135857 44.612 43.071 175.30 21.5768 2.26790 37.199417692 36.587308722 1292.66 201.87 59338.22 62972.55 63562.29 55555.51 54648.29 85.282 25.2658 41.179 263.8 16019.16 17823.972 11.44 58.75 13.67 22.061 126457 139035 18304.39 22528.92 16.6699677 9531.48 1559.3 14.68 14.48 3.72571 22577.88 7.54 16190.133 17705.756 5022 6.33617 11196.8 3903 145.78 146.707 4.457 4.409 54.41 4.145 3.43803 971.785 880.51 282841 676.53 512.546 1.25 335.53 239.446 235.05 189.28942 26.766308 22590000 2.08 36.12 193.87 180.32 1.366 732078 29.246 34193 11771.22 68.71 68.86 33.85 7.43 104.36 101.40844005 98.4589975 100.26 30311842 95.28 0.62459 114660.49 83.36 2.182 76.5668259 38315.1 75.672 31845 65.25 7.57 5.58648 113.89 323.77 18.51 59.82 100.21 79.41 75.48 79.28 75.64 17.16 349.13 5.41 1104.95 18480000 20.83 287.49 9.08 658.28 92 167 188 261 289 3209986 17799.501 111.7 18404.785 41210.06 46.53244121 44.181 42.943 178.26 21.8505 2.27071 37.202661409 35.537095139 1345.7 205.306 61195.82 65210.1 65913.62 57010.05 56856.03 87.054 25.6014 41.745 266.66 17776.419 18169.531 11.85 58.87 14.69 21.995 126189 141015 19269.37 23481.95 15.9998989 10133.13 1558.7 14.71 15 3.58111 23251.34 7.382 17220.318 18105.09 4952 6.10032 11166.1 3868 147.738 148.201 4.354 4.287 54.77 4.122 3.43041 OpenBenchmarking.org
Timed Linux Kernel Compilation Build: allmodconfig OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.8 Build: allmodconfig 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 200 400 600 800 1000 971.79 979.12
OpenRadioss Model: Chrysler Neon 1M OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Chrysler Neon 1M 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 200 400 600 800 1000 880.51 913.48
BRL-CAD VGR Performance Metric OpenBenchmarking.org VGR Performance Metric, More Is Better BRL-CAD 7.38.2 VGR Performance Metric 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 60K 120K 180K 240K 300K 282841 280941 1. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6
Blender Blend File: Barbershop - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Barbershop - Compute: CPU-Only 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 150 300 450 600 750 676.26 676.53
Timed Node.js Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 21.7.2 Time To Compile 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 110 220 330 440 550 512.55 515.33
Llama.cpp Model: llama-2-70b-chat.Q5_0.gguf OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b1808 Model: llama-2-70b-chat.Q5_0.gguf 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.2813 0.5626 0.8439 1.1252 1.4065 1.25 1.22 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
OpenRadioss Model: INIVOL and Fluid Structure Interaction Drop Container OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: INIVOL and Fluid Structure Interaction Drop Container 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 70 140 210 280 350 335.53 338.34
Timed Godot Game Engine Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Godot Game Engine Compilation 4.0 Time To Compile 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 50 100 150 200 250 239.45 240.33
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Pabellon Barcelona - Compute: CPU-Only 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 50 100 150 200 250 233.85 235.05
OpenFOAM Input: drivaerFastback, Small Mesh Size - Execution Time OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Execution Time 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 40 80 120 160 200 189.29 196.38 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
OpenFOAM Input: drivaerFastback, Small Mesh Size - Mesh Time OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Mesh Time 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 7 14 21 28 35 26.77 28.68 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfoamToVTK -ldynamicMesh -llagrangian -lgenericPatchFields -lfileFormats -lOpenFOAM -ldl -lm
Quicksilver Input: CORAL2 P2 OpenBenchmarking.org Figure Of Merit, More Is Better Quicksilver 20230818 Input: CORAL2 P2 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5M 10M 15M 20M 25M 22590000 22580000 1. (CXX) g++ options: -fopenmp -O3 -march=native
Llamafile Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPU OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.7 Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.468 0.936 1.404 1.872 2.34 2.08 2.02
TensorFlow Device: CPU - Batch Size: 64 - Model: ResNet-50 OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: ResNet-50 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 8 16 24 32 40 36.12 35.44
Blender Blend File: Classroom - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Classroom - Compute: CPU-Only 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 40 80 120 160 200 193.27 193.87
OpenRadioss Model: Bird Strike on Windshield OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Bird Strike on Windshield 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 40 80 120 160 200 180.32 182.19
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.3116 0.6232 0.9348 1.2464 1.558 1.366 1.385 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Only OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 160K 320K 480K 640K 800K 732078 721872 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 7 14 21 28 35 29.25 29.47 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Write OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 7K 14K 21K 28K 35K 34193 33933 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
NAS Parallel Benchmarks Test / Class: SP.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 3K 6K 9K 12K 15K 11771.22 11331.51 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
FFmpeg Encoder: libx265 - Scenario: Platform OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Platform 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 15 30 45 60 75 68.71 68.30 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
FFmpeg Encoder: libx265 - Scenario: Video On Demand OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Video On Demand 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 15 30 45 60 75 68.86 68.27 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
FFmpeg Encoder: libx265 - Scenario: Upload OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Upload 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 8 16 24 32 40 33.85 33.66 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
Llamafile Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPU OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.7 Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 2 4 6 8 10 7.43 7.21
OpenRadioss Model: Bumper Beam OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Bumper Beam 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 104.36 105.47
SPECFEM3D Model: Layered Halfspace OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.1.1 Model: Layered Halfspace 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 101.41 102.83 1. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
SPECFEM3D Model: Water-layered Halfspace OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.1.1 Model: Water-layered Halfspace 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 98.46 100.60 1. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Blender Blend File: Junkshop - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Junkshop - Compute: CPU-Only 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 20 40 60 80 100 99.78 100.26
Stockfish Chess Benchmark OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 16.1 Chess Benchmark 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 6M 12M 18M 24M 30M 30311842 30072838 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver
Blender Blend File: Fishy Cat - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.1 Blend File: Fishy Cat - Compute: CPU-Only 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 20 40 60 80 100 94.72 95.28
NAMD Input: STMV with 1,066,628 Atoms OpenBenchmarking.org ns/day, More Is Better NAMD 3.0b6 Input: STMV with 1,066,628 Atoms 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.1405 0.281 0.4215 0.562 0.7025 0.62459 0.62007
nginx Connections: 1000 OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.23.2 Connections: 1000 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 20K 40K 60K 80K 100K 114755.20 114660.49 1. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenRadioss Model: Rubber O-Ring Seal Installation OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Rubber O-Ring Seal Installation 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 83.36 83.39
GROMACS Implementation: MPI CPU - Input: water_GMX50_bare OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.491 0.982 1.473 1.964 2.455 2.182 2.151 1. (CXX) g++ options: -O3 -lm
Xcompact3d Incompact3d Input: input.i3d 193 Cells Per Direction OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 76.57 79.43 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
NAS Parallel Benchmarks Test / Class: BT.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 8K 16K 24K 32K 40K 38315.10 36869.25 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
Timed Linux Kernel Compilation Build: defconfig OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.8 Build: defconfig 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 75.67 76.58
Chaos Group V-RAY Mode: CPU OpenBenchmarking.org vsamples, More Is Better Chaos Group V-RAY 6.0 Mode: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 7K 14K 21K 28K 35K 31845 31844
OpenRadioss Model: Cell Phone Drop Test OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Cell Phone Drop Test 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 15 30 45 60 75 65.18 65.25
Llama.cpp Model: llama-2-13b.Q4_0.gguf OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b1808 Model: llama-2-13b.Q4_0.gguf 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 2 4 6 8 10 7.57 7.30 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
OSPRay Benchmark: gravity_spheres_volume/dim_512/ao/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPRay 3.1 Benchmark: gravity_spheres_volume/dim_512/ao/real_time 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 1.257 2.514 3.771 5.028 6.285 5.58648 5.49836
TensorFlow Device: CPU - Batch Size: 64 - Model: GoogLeNet OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: GoogLeNet 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 30 60 90 120 150 113.89 112.13
OpenVINO Model: Face Detection FP16-INT8 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Face Detection FP16-INT8 - Device: CPU 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 70 140 210 280 350 322.63 323.77 MIN: 296.85 / MAX: 328.51 MIN: 308.71 / MAX: 330.44 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Face Detection FP16-INT8 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Face Detection FP16-INT8 - Device: CPU 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 5 10 15 20 25 18.57 18.51 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Machine Translation EN To DE FP16 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Machine Translation EN To DE FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 14 28 42 56 70 59.82 60.53 MIN: 48.42 / MAX: 67.39 MIN: 33.29 / MAX: 82.21 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Machine Translation EN To DE FP16 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Machine Translation EN To DE FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 100.21 99.05 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Detection FP16 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 79.41 80.58 MIN: 56.3 / MAX: 102.74 MIN: 60.22 / MAX: 102.9 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Detection FP16 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 75.48 74.40 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Detection FP32 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Detection FP32 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 79.28 80.04 MIN: 60.22 / MAX: 100.15 MIN: 61.76 / MAX: 107.58 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Detection FP32 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Detection FP32 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 75.64 74.91 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Road Segmentation ADAS FP16-INT8 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4 8 12 16 20 17.16 17.32 MIN: 13.76 / MAX: 23 MIN: 8.73 / MAX: 23.67 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Road Segmentation ADAS FP16-INT8 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 80 160 240 320 400 349.13 345.80 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Vehicle Bike Detection FP16 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Person Vehicle Bike Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 1.233 2.466 3.699 4.932 6.165 5.41 5.48 MIN: 3.92 / MAX: 8.18 MIN: 3.89 / MAX: 9.6 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Person Vehicle Bike Detection FP16 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Person Vehicle Bike Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 200 400 600 800 1000 1104.95 1090.85 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Quicksilver Input: CORAL2 P1 OpenBenchmarking.org Figure Of Merit, More Is Better Quicksilver 20230818 Input: CORAL2 P1 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 4M 8M 12M 16M 20M 18550000 18480000 1. (CXX) g++ options: -fopenmp -O3 -march=native
OpenVINO Model: Road Segmentation ADAS FP16 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5 10 15 20 25 20.83 21.27 MIN: 12.6 / MAX: 29.81 MIN: 9.77 / MAX: 37.47 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Road Segmentation ADAS FP16 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Road Segmentation ADAS FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 60 120 180 240 300 287.49 281.64 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Vehicle Detection FP16 - Device: CPU OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2024.0 Model: Vehicle Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 3 6 9 12 15 9.08 9.26 MIN: 4.23 / MAX: 17.05 MIN: 4.6 / MAX: 15.49 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenVINO Model: Vehicle Detection FP16 - Device: CPU OpenBenchmarking.org FPS, More Is Better OpenVINO 2024.0 Model: Vehicle Detection FP16 - Device: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 140 280 420 560 700 658.28 645.63 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
GraphicsMagick Operation: Noise-Gaussian OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Noise-Gaussian 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 92 91 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp
GraphicsMagick Operation: Rotate OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Rotate 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 40 80 120 160 200 167 165 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp
GraphicsMagick Operation: Swirl OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Swirl 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 40 80 120 160 200 188 187 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp
GraphicsMagick Operation: HWB Color Space OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: HWB Color Space 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 60 120 180 240 300 261 259 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp
GraphicsMagick Operation: Resizing OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Resizing 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 60 120 180 240 300 289 289 1. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp
RocksDB Test: Read While Writing OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read While Writing 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 700K 1400K 2100K 2800K 3500K 3209986 3174105 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
MBW Test: Memory Copy, Fixed Block Size - Array Size: 8192 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy, Fixed Block Size - Array Size: 8192 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 17799.50 16405.97 1. (CC) gcc options: -O3 -march=native
libxsmm M N K: 32 OpenBenchmarking.org GFLOPS/s, More Is Better libxsmm 2-1.17-3645 M N K: 32 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 30 60 90 120 150 111.7 107.1 1. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -pedantic -O2 -fopenmp -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -march=core-avx2
MBW Test: Memory Copy - Array Size: 8192 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy - Array Size: 8192 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 18404.79 17709.53 1. (CC) gcc options: -O3 -march=native
NAS Parallel Benchmarks Test / Class: LU.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 9K 18K 27K 36K 45K 41210.06 41111.39 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
SPECFEM3D Model: Homogeneous Halfspace OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.1.1 Model: Homogeneous Halfspace 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 11 22 33 44 55 46.53 47.02 1. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
Timed PHP Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed PHP Compilation 8.3.4 Time To Compile 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 10 20 30 40 50 44.18 44.61
libavif avifenc Encoder Speed: 2 OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 2 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 10 20 30 40 50 42.94 43.07 1. (CXX) g++ options: -O3 -fPIC -lm
FFmpeg Encoder: libx265 - Scenario: Live OpenBenchmarking.org FPS, More Is Better FFmpeg 7.0 Encoder: libx265 - Scenario: Live 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 40 80 120 160 200 178.26 175.30 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
Embree Binary: Pathtracer ISPC - Model: Asian Dragon Obj OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon Obj 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5 10 15 20 25 21.85 21.58 MIN: 21.73 / MAX: 22.2 MIN: 21.45 / MAX: 21.94
NAMD Input: ATPase with 327,506 Atoms OpenBenchmarking.org ns/day, More Is Better NAMD 3.0b6 Input: ATPase with 327,506 Atoms 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.5109 1.0218 1.5327 2.0436 2.5545 2.27071 2.26790
SPECFEM3D Model: Mount St. Helens OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.1.1 Model: Mount St. Helens 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 9 18 27 36 45 37.20 37.20 1. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
SPECFEM3D Model: Tomographic Model OpenBenchmarking.org Seconds, Fewer Is Better SPECFEM3D 4.1.1 Model: Tomographic Model 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 8 16 24 32 40 35.54 36.59 1. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
NAS Parallel Benchmarks Test / Class: IS.D OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 300 600 900 1200 1500 1345.70 1292.66 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
JPEG-XL Decoding libjxl CPU Threads: All OpenBenchmarking.org MP/s, More Is Better JPEG-XL Decoding libjxl 0.10.1 CPU Threads: All 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 50 100 150 200 250 205.31 201.87
RAMspeed SMP Type: Average - Benchmark: Integer OpenBenchmarking.org MB/s, More Is Better RAMspeed SMP 3.5.0 Type: Average - Benchmark: Integer 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 13K 26K 39K 52K 65K 61195.82 59338.22 1. (CC) gcc options: -O3 -march=native
RAMspeed SMP Type: Copy - Benchmark: Integer OpenBenchmarking.org MB/s, More Is Better RAMspeed SMP 3.5.0 Type: Copy - Benchmark: Integer 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 14K 28K 42K 56K 70K 65210.10 62972.55 1. (CC) gcc options: -O3 -march=native
RAMspeed SMP Type: Scale - Benchmark: Integer OpenBenchmarking.org MB/s, More Is Better RAMspeed SMP 3.5.0 Type: Scale - Benchmark: Integer 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 14K 28K 42K 56K 70K 65913.62 63562.29 1. (CC) gcc options: -O3 -march=native
RAMspeed SMP Type: Triad - Benchmark: Integer OpenBenchmarking.org MB/s, More Is Better RAMspeed SMP 3.5.0 Type: Triad - Benchmark: Integer 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 12K 24K 36K 48K 60K 57010.05 55555.51 1. (CC) gcc options: -O3 -march=native
RAMspeed SMP Type: Add - Benchmark: Integer OpenBenchmarking.org MB/s, More Is Better RAMspeed SMP 3.5.0 Type: Add - Benchmark: Integer 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 12K 24K 36K 48K 60K 56856.03 54648.29 1. (CC) gcc options: -O3 -march=native
JPEG-XL Decoding libjxl CPU Threads: 1 OpenBenchmarking.org MP/s, More Is Better JPEG-XL Decoding libjxl 0.10.1 CPU Threads: 1 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 87.05 85.28
Embree Binary: Pathtracer ISPC - Model: Asian Dragon OpenBenchmarking.org Frames Per Second, More Is Better Embree 4.3 Binary: Pathtracer ISPC - Model: Asian Dragon 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 6 12 18 24 30 25.60 25.27 MIN: 25.46 / MAX: 25.94 MIN: 25.12 / MAX: 25.79
JPEG-XL libjxl Input: PNG - Quality: 90 OpenBenchmarking.org MP/s, More Is Better JPEG-XL libjxl 0.10.1 Input: PNG - Quality: 90 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 10 20 30 40 50 41.75 41.18 1. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm
TensorFlow Device: CPU - Batch Size: 64 - Model: AlexNet OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 64 - Model: AlexNet 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 60 120 180 240 300 266.66 263.80
MBW Test: Memory Copy, Fixed Block Size - Array Size: 4096 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy, Fixed Block Size - Array Size: 4096 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 17776.42 16019.16 1. (CC) gcc options: -O3 -march=native
MBW Test: Memory Copy - Array Size: 4096 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy - Array Size: 4096 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 18169.53 17823.97 1. (CC) gcc options: -O3 -march=native
Llamafile Test: llava-v1.5-7b-q4 - Acceleration: CPU OpenBenchmarking.org Tokens Per Second, More Is Better Llamafile 0.7 Test: llava-v1.5-7b-q4 - Acceleration: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 3 6 9 12 15 11.85 11.44
PyTorch Device: CPU - Batch Size: 1 - Model: ResNet-50 OpenBenchmarking.org batches/sec, More Is Better PyTorch 2.2.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 13 26 39 52 65 58.87 58.75 MIN: 54.65 / MAX: 59.36 MIN: 53.89 / MAX: 59.52
Llama.cpp Model: llama-2-7b.Q4_0.gguf OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b1808 Model: llama-2-7b.Q4_0.gguf 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4 8 12 16 20 14.69 13.67 1. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
Timed Mesa Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Mesa Compilation 24.0 Time To Compile 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5 10 15 20 25 22.00 22.06
7-Zip Compression Test: Decompression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Decompression Rating 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 30K 60K 90K 120K 150K 126457 126189 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
7-Zip Compression Test: Compression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Compression Rating 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 30K 60K 90K 120K 150K 141015 139035 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
NAS Parallel Benchmarks Test / Class: SP.B OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 19269.37 18304.39 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
NAS Parallel Benchmarks Test / Class: FT.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5K 10K 15K 20K 25K 23481.95 22528.92 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
Xcompact3d Incompact3d Input: input.i3d 129 Cells Per Direction OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 129 Cells Per Direction 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4 8 12 16 20 16.00 16.67 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz
NAS Parallel Benchmarks Test / Class: CG.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 2K 4K 6K 8K 10K 10133.13 9531.48 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
srsRAN Project Test: PUSCH Processor Benchmark, Throughput Total OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Total 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 300 600 900 1200 1500 1559.3 1558.7 MIN: 954.9 MIN: 954.9 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl
TensorFlow Device: CPU - Batch Size: 1 - Model: ResNet-50 OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: ResNet-50 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4 8 12 16 20 14.71 14.68
TensorFlow Device: CPU - Batch Size: 1 - Model: AlexNet OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: AlexNet 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4 8 12 16 20 15.00 14.48
oneDNN Harness: IP Shapes 3D - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.4 Harness: IP Shapes 3D - Engine: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.8383 1.6766 2.5149 3.3532 4.1915 3.58111 3.72571 MIN: 3.54 MIN: 3.69 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl
NAS Parallel Benchmarks Test / Class: MG.C OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 5K 10K 15K 20K 25K 23251.34 22577.88 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.6
RNNoise Input: 26 Minute Long Talking Sample OpenBenchmarking.org Seconds, Fewer Is Better RNNoise 0.2 Input: 26 Minute Long Talking Sample 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 2 4 6 8 10 7.382 7.540 1. (CC) gcc options: -O2 -pedantic -fvisibility=hidden
MBW Test: Memory Copy, Fixed Block Size - Array Size: 1024 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy, Fixed Block Size - Array Size: 1024 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 17220.32 16190.13 1. (CC) gcc options: -O3 -march=native
MBW Test: Memory Copy - Array Size: 1024 MiB OpenBenchmarking.org MiB/s, More Is Better MBW 2018-09-08 Test: Memory Copy - Array Size: 1024 MiB 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 4K 8K 12K 16K 20K 18105.09 17705.76 1. (CC) gcc options: -O3 -march=native
Google Draco Model: Church Facade OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.6 Model: Church Facade 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 1100 2200 3300 4400 5500 4952 5022 1. (CXX) g++ options: -O3
oneDNN Harness: Convolution Batch Shapes Auto - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.4 Harness: Convolution Batch Shapes Auto - Engine: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 2 4 6 8 10 6.10032 6.33617 MIN: 5.99 MIN: 6.24 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl
srsRAN Project Test: PDSCH Processor Benchmark, Throughput Total OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Total 2 x 32GB DDR5-4800 2 x 32GB DDR5-5200 2K 4K 6K 8K 10K 11196.8 11166.1 1. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl
Google Draco Model: Lion OpenBenchmarking.org ms, Fewer Is Better Google Draco 1.5.6 Model: Lion 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 800 1600 2400 3200 4000 3868 3903 1. (CXX) g++ options: -O3
SVT-AV1 Encoder Mode: Preset 12 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 12 - Input: Bosphorus 4K 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 30 60 90 120 150 147.74 145.78 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
SVT-AV1 Encoder Mode: Preset 13 - Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.0 Encoder Mode: Preset 13 - Input: Bosphorus 4K 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 30 60 90 120 150 148.20 146.71 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
libavif avifenc Encoder Speed: 6 OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 6 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 1.0028 2.0056 3.0084 4.0112 5.014 4.354 4.457 1. (CXX) g++ options: -O3 -fPIC -lm
WavPack Audio Encoding WAV To WavPack OpenBenchmarking.org Seconds, Fewer Is Better WavPack Audio Encoding 5.7 WAV To WavPack 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.992 1.984 2.976 3.968 4.96 4.287 4.409
TensorFlow Device: CPU - Batch Size: 1 - Model: GoogLeNet OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 1 - Model: GoogLeNet 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 12 24 36 48 60 54.77 54.41
libavif avifenc Encoder Speed: 10, Lossless OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 10, Lossless 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.9326 1.8652 2.7978 3.7304 4.663 4.122 4.145 1. (CXX) g++ options: -O3 -fPIC -lm
oneDNN Harness: Deconvolution Batch shapes_3d - Engine: CPU OpenBenchmarking.org ms, Fewer Is Better oneDNN 3.4 Harness: Deconvolution Batch shapes_3d - Engine: CPU 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 0.7736 1.5472 2.3208 3.0944 3.868 3.43041 3.43803 MIN: 3.21 MIN: 3.21 1. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl
Geometric Mean Of All Test Results Result Composite - AMD Ryzen 9 7900 DDR5-5200 ECC Memory OpenBenchmarking.org Geometric Mean, More Is Better Geometric Mean Of All Test Results Result Composite - AMD Ryzen 9 7900 DDR5-5200 ECC Memory 2 x 32GB DDR5-5200 2 x 32GB DDR5-4800 20 40 60 80 100 90.27 88.66
Phoronix Test Suite v10.8.4