ff

AMD Ryzen Threadripper 3970X 32-Core testing with a ASUS ROG ZENITH II EXTREME (1802 BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 22.04 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2401113-NE-FF610899407
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results

Limit displaying results to tests within:

CPU Massive 2 Tests
HPC - High Performance Computing 3 Tests
Machine Learning 3 Tests
Python Tests 2 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
a
January 28
  2 Hours, 56 Minutes
b
January 29
  1 Hour
Invert Hiding All Results Option
  1 Hour, 58 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

a

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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 Notes: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107a
Python Notes: Python 3.10.12
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

b

Processor: AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH II EXTREME (1802 BIOS), Chipset: AMD Starship/Matisse, Memory: 4 x 16 GB DRAM-3600MT/s Corsair CMT64GX4M4Z3600C16, Disk: Samsung SSD 980 PRO 500GB, Graphics: AMD Radeon RX 5700 8GB (1750/875MHz), Audio: AMD Navi 10 HDMI Audio, Monitor: ASUS VP28U, Network: Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200

OS: Ubuntu 22.04, Kernel: 6.2.0-39-generic (x86_64), Desktop: GNOME Shell 42.2, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.0.1 (LLVM 13.0.1 DRM 3.49), Vulkan: 1.2.204, Compiler: GCC 11.4.0, File-System: ext4, Screen Resolution: 3840x2160

ffOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads)ASUS ROG ZENITH II EXTREME (1802 BIOS)AMD Starship/Matisse4 x 16 GB DRAM-3600MT/s Corsair CMT64GX4M4Z3600C16Samsung SSD 980 PRO 500GBAMD Radeon RX 5700 8GB (1750/875MHz)AMD Navi 10 HDMI AudioASUS VP28UAquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200Ubuntu 22.046.2.0-39-generic (x86_64)GNOME Shell 42.2X Server + Wayland4.6 Mesa 22.0.1 (LLVM 13.0.1 DRM 3.49)1.2.204GCC 11.4.0ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLVulkanCompilerFile-SystemScreen ResolutionFf BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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 - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107a- Python 3.10.12- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

ffquicksilver: CTS2quicksilver: CORAL2 P1quicksilver: CORAL2 P2cachebench: Readcachebench: Writecachebench: Read / Modify / Writerav1e: 1rav1e: 5rav1e: 6rav1e: 10y-cruncher: 1By-cruncher: 500Mpytorch: CPU - 1 - ResNet-50pytorch: CPU - 1 - ResNet-152pytorch: CPU - 16 - ResNet-50pytorch: CPU - 16 - ResNet-152pytorch: CPU - 1 - Efficientnet_v2_lpytorch: CPU - 16 - Efficientnet_v2_ltensorflow: CPU - 1 - VGG-16tensorflow: CPU - 1 - AlexNettensorflow: CPU - 16 - VGG-16tensorflow: CPU - 16 - AlexNettensorflow: CPU - 1 - GoogLeNettensorflow: CPU - 1 - ResNet-50tensorflow: CPU - 16 - GoogLeNettensorflow: CPU - 16 - ResNet-50speedb: Rand Fillspeedb: Rand Readspeedb: Update Randspeedb: Seq Fillspeedb: Rand Fill Syncspeedb: Read While Writingspeedb: Read Rand Write Randllama-cpp: llama-2-7b.Q4_0.ggufllama-cpp: llama-2-13b.Q4_0.ggufllama-cpp: llama-2-70b-chat.Q5_0.ggufab20786667219600002895000011066.95543761568.843405119470.6580170.8362.8133.7338.63016.3737.92535.7913.9530.6811.747.905.992.075.356.9061.419.236.3742.9012.5425245512949577624171025412256187991264242214419.3610.631.8620850000219900002895000011033.0293161337.271973118721.3953370.8392.8093.788.86916.3947.98635.4314.1931.1611.887.896.022.015.356.8160.429.236.443.2912.5925225113147573424095725435356138052776240310319.3210.641.86OpenBenchmarking.org

Quicksilver

Quicksilver is a proxy application that represents some elements of the Mercury workload by solving a simplified dynamic Monte Carlo particle transport problem. Quicksilver is developed by Lawrence Livermore National Laboratory (LLNL) and this test profile currently makes use of the OpenMP CPU threaded code path. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CTS2ab4M8M12M16M20MSE +/- 37564.76, N = 320786667208500001. (CXX) g++ options: -fopenmp -O3 -march=native
OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CTS2ab4M8M12M16M20MMin: 20720000 / Avg: 20786666.67 / Max: 208500001. (CXX) g++ options: -fopenmp -O3 -march=native

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P1ab5M10M15M20M25MSE +/- 0.00, N = 321960000219900001. (CXX) g++ options: -fopenmp -O3 -march=native
OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P1ab4M8M12M16M20MMin: 21960000 / Avg: 21960000 / Max: 219600001. (CXX) g++ options: -fopenmp -O3 -march=native

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P2ab6M12M18M24M30MSE +/- 37859.39, N = 328950000289500001. (CXX) g++ options: -fopenmp -O3 -march=native
OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P2ab5M10M15M20M25MMin: 28890000 / Avg: 28950000 / Max: 290200001. (CXX) g++ options: -fopenmp -O3 -march=native

CacheBench

This is a performance test of CacheBench, which is part of LLCbench. CacheBench is designed to test the memory and cache bandwidth performance Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Readab2K4K6K8K10KSE +/- 33.97, N = 311066.9611033.03MIN: 10956.53 / MAX: 11112.67MIN: 11002.15 / MAX: 11058.911. (CC) gcc options: -O3 -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Readab2K4K6K8K10KMin: 10999.03 / Avg: 11066.96 / Max: 11102.031. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Writeab13K26K39K52K65KSE +/- 29.06, N = 361568.8461337.27MIN: 40788.42 / MAX: 66208.33MIN: 41835.46 / MAX: 65734.931. (CC) gcc options: -O3 -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Writeab11K22K33K44K55KMin: 61525.83 / Avg: 61568.84 / Max: 61624.21. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / Writeab30K60K90K120K150KSE +/- 81.20, N = 3119470.66118721.40MIN: 97982.3 / MAX: 130920.18MIN: 99330.52 / MAX: 130732.061. (CC) gcc options: -O3 -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / Writeab20K40K60K80K100KMin: 119383.08 / Avg: 119470.66 / Max: 119632.91. (CC) gcc options: -O3 -lrt

rav1e

Xiph rav1e is a Rust-written AV1 video encoder that claims to be the fastest and safest AV1 encoder. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 1ab0.18880.37760.56640.75520.944SE +/- 0.001, N = 30.8360.839
OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 1ab246810Min: 0.84 / Avg: 0.84 / Max: 0.84

OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 5ab0.63291.26581.89872.53163.1645SE +/- 0.005, N = 32.8132.809
OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 5ab246810Min: 2.8 / Avg: 2.81 / Max: 2.82

OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 6ab0.85051.7012.55153.4024.2525SE +/- 0.017, N = 33.7333.780
OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 6ab246810Min: 3.72 / Avg: 3.73 / Max: 3.77

OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 10ab246810SE +/- 0.075, N = 38.6308.869
OpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.7Speed: 10ab3691215Min: 8.48 / Avg: 8.63 / Max: 8.71

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1Bab48121620SE +/- 0.03, N = 316.3716.39
OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1Bab48121620Min: 16.33 / Avg: 16.37 / Max: 16.43

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500Mab246810SE +/- 0.007, N = 37.9257.986
OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500Mab3691215Min: 7.91 / Avg: 7.93 / Max: 7.94

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Currently this test profile is catered to CPU-based testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-50ab816243240SE +/- 0.14, N = 335.7935.43MIN: 35 / MAX: 36.54MIN: 34.31 / MAX: 36.32
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-50ab816243240Min: 35.55 / Avg: 35.79 / Max: 36.05

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152ab48121620SE +/- 0.08, N = 313.9514.19MIN: 13.59 / MAX: 14.24MIN: 14.01 / MAX: 14.36
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152ab48121620Min: 13.86 / Avg: 13.95 / Max: 14.1

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-50ab714212835SE +/- 0.17, N = 330.6831.16MIN: 30 / MAX: 31.31MIN: 30.47 / MAX: 31.36
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-50ab714212835Min: 30.46 / Avg: 30.68 / Max: 31.01

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-152ab3691215SE +/- 0.05, N = 311.7411.88MIN: 11.43 / MAX: 11.97MIN: 11.8 / MAX: 11.95
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: ResNet-152ab3691215Min: 11.69 / Avg: 11.74 / Max: 11.83

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lab246810SE +/- 0.01, N = 37.907.89MIN: 7.78 / MAX: 8.01MIN: 7.8 / MAX: 7.97
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lab3691215Min: 7.88 / Avg: 7.9 / Max: 7.92

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lab246810SE +/- 0.02, N = 35.996.02MIN: 5.9 / MAX: 6.06MIN: 5.98 / MAX: 6.06
OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lab246810Min: 5.96 / Avg: 5.99 / Max: 6.02

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: VGG-16ab0.46580.93161.39741.86322.329SE +/- 0.01, N = 32.072.01
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: VGG-16ab246810Min: 2.05 / Avg: 2.07 / Max: 2.08

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: AlexNetab1.20382.40763.61144.81526.019SE +/- 0.00, N = 35.355.35
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: AlexNetab246810Min: 5.34 / Avg: 5.35 / Max: 5.35

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: VGG-16ab246810SE +/- 0.01, N = 36.906.81
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: VGG-16ab3691215Min: 6.89 / Avg: 6.9 / Max: 6.91

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: AlexNetab1428425670SE +/- 0.13, N = 361.4160.42
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: AlexNetab1224364860Min: 61.22 / Avg: 61.41 / Max: 61.66

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: GoogLeNetab3691215SE +/- 0.02, N = 39.239.23
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: GoogLeNetab3691215Min: 9.2 / Avg: 9.23 / Max: 9.26

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: ResNet-50ab246810SE +/- 0.01, N = 36.376.40
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 1 - Model: ResNet-50ab3691215Min: 6.36 / Avg: 6.37 / Max: 6.4

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: GoogLeNetab1020304050SE +/- 0.16, N = 342.9043.29
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: GoogLeNetab918273645Min: 42.74 / Avg: 42.9 / Max: 43.22

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: ResNet-50ab3691215SE +/- 0.08, N = 312.5412.59
OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: ResNet-50ab48121620Min: 12.37 / Avg: 12.54 / Max: 12.63

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fillab50K100K150K200K250KSE +/- 91.82, N = 32524552522511. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fillab40K80K120K160K200KMin: 252313 / Avg: 252455.33 / Max: 2526271. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Readab30M60M90M120M150MSE +/- 1683044.27, N = 31294957761314757341. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Readab20M40M60M80M100MMin: 127566334 / Avg: 129495776 / Max: 1328491901. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update Randomab50K100K150K200K250KSE +/- 580.55, N = 32417102409571. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update Randomab40K80K120K160K200KMin: 240596 / Avg: 241709.67 / Max: 2425511. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Sequential Fillab50K100K150K200K250KSE +/- 336.51, N = 32541222543531. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Sequential Fillab40K80K120K160K200KMin: 253754 / Avg: 254122 / Max: 2547941. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fill Syncab12002400360048006000SE +/- 49.65, N = 3561856131. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fill Syncab10002000300040005000Min: 5532 / Avg: 5617.67 / Max: 57041. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While Writingab2M4M6M8M10MSE +/- 69727.75, N = 3799126480527761. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While Writingab1.4M2.8M4.2M5.6M7MMin: 7867281 / Avg: 7991264 / Max: 81085461. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write Randomab500K1000K1500K2000K2500KSE +/- 3556.84, N = 3242214424031031. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write Randomab400K800K1200K1600K2000KMin: 2415493 / Avg: 2422144 / Max: 24276551. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Llama.cpp

Llama.cpp is a port of Facebook's LLaMA model in C/C++ developed by Georgi Gerganov. Llama.cpp allows the inference of LLaMA and other supported models in C/C++. For CPU inference Llama.cpp supports AVX2/AVX-512, ARM NEON, and other modern ISAs along with features like OpenBLAS usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-7b.Q4_0.ggufab510152025SE +/- 0.07, N = 319.3619.321. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-7b.Q4_0.ggufab510152025Min: 19.29 / Avg: 19.36 / Max: 19.511. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-13b.Q4_0.ggufab3691215SE +/- 0.00, N = 310.6310.641. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-13b.Q4_0.ggufab3691215Min: 10.63 / Avg: 10.63 / Max: 10.631. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-70b-chat.Q5_0.ggufab0.41850.8371.25551.6742.0925SE +/- 0.00, N = 31.861.861. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas
OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b1808Model: llama-2-70b-chat.Q5_0.ggufab246810Min: 1.86 / Avg: 1.86 / Max: 1.871. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

a

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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 Notes: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107a
Python Notes: Python 3.10.12
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 28 January 2024 20:40 by user phoronix.

b

Processor: AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads), Motherboard: ASUS ROG ZENITH II EXTREME (1802 BIOS), Chipset: AMD Starship/Matisse, Memory: 4 x 16 GB DRAM-3600MT/s Corsair CMT64GX4M4Z3600C16, Disk: Samsung SSD 980 PRO 500GB, Graphics: AMD Radeon RX 5700 8GB (1750/875MHz), Audio: AMD Navi 10 HDMI Audio, Monitor: ASUS VP28U, Network: Aquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200

OS: Ubuntu 22.04, Kernel: 6.2.0-39-generic (x86_64), Desktop: GNOME Shell 42.2, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.0.1 (LLVM 13.0.1 DRM 3.49), Vulkan: 1.2.204, Compiler: GCC 11.4.0, File-System: ext4, Screen Resolution: 3840x2160

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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 Notes: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x830107a
Python Notes: Python 3.10.12
Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 29 January 2024 00:22 by user phoronix.