New Kernels EPYC

2 x AMD EPYC 9684X 96-Core testing with a AMD Titanite_4G (RTI1007B BIOS) and ASPEED on Ubuntu 23.10 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 2401268-NE-NEWKERNEL89
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
Show Result Confidence Charts

Limit displaying results to tests within:

Timed Code Compilation 5 Tests
C/C++ Compiler Tests 3 Tests
Compression Tests 2 Tests
CPU Massive 8 Tests
Creator Workloads 3 Tests
Database Test Suite 2 Tests
Game Development 2 Tests
HPC - High Performance Computing 5 Tests
Common Kernel Benchmarks 3 Tests
Large Language Models 2 Tests
Machine Learning 5 Tests
Multi-Core 9 Tests
Programmer / Developer System Benchmarks 5 Tests
Python Tests 6 Tests
Server 3 Tests
Server CPU Tests 4 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
Linux 6.5
January 24
  8 Hours, 37 Minutes
Linux 6.6
January 25
  8 Hours, 29 Minutes
Linux 6.7
January 25
  9 Hours, 3 Minutes
Linux 6.8 26 Jan
January 26
  9 Hours, 54 Minutes
Invert Hiding All Results Option
  9 Hours, 1 Minute

Only show results where is faster than
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):


New Kernels EPYCOpenBenchmarking.orgPhoronix Test Suite2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads)AMD Titanite_4G (RTI1007B BIOS)AMD Device 14a41520GB3201GB Micron_7450_MTFDKCB3T2TFSASPEEDBroadcom NetXtreme BCM5720 PCIeUbuntu 23.106.5.0-060500-generic (x86_64)6.6.0-060600-generic (x86_64)6.7.0-060700-generic (x86_64)6.8.0-060800rc1daily20240126-generic (x86_64)GCC 13.2.0ext4800x600ProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelsCompilerFile-SystemScreen ResolutionNew Kernels EPYC 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,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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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-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 performance (Boost: Enabled) - CPU Microcode: 0xa10113e - Python 3.11.6- Linux 6.5: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - Linux 6.6: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - Linux 6.7: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - Linux 6.8 26 Jan: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + 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 Enhanced / Automatic IBRS IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 JanResult OverviewPhoronix Test Suite100%108%116%124%132%PyTorchLlama.cppRocksDBTimed Linux Kernel CompilationStress-NGSpeedbPostgreSQLTimed Gem5 CompilationTimed Apache CompilationTensorFlowTimed Node.js CompilationTimed Godot Game Engine CompilationY-Cruncher7-Zip CompressionLlamafileBlenderOpenVINOLZ4 CompressionCacheBench

New Kernels EPYCcachebench: Readcachebench: Writecachebench: Read / Modify / Writecompress-lz4: 1 - Compression Speedcompress-lz4: 1 - Decompression Speedcompress-lz4: 3 - Compression Speedcompress-lz4: 3 - Decompression Speedcompress-lz4: 9 - Compression Speedcompress-lz4: 9 - Decompression Speedcompress-7zip: Compression Ratingcompress-7zip: Decompression Ratingbuild-apache: Time To Compilebuild-gem5: Time To Compilebuild-godot: Time To Compilebuild-linux-kernel: defconfigbuild-linux-kernel: allmodconfigbuild-nodejs: Time To Compiley-cruncher: 1By-cruncher: 500Mpgbench: 100 - 1000 - Read Onlypgbench: 100 - 1000 - Read Only - Average Latencypgbench: 100 - 1000 - Read Writepgbench: 100 - 1000 - Read Write - Average Latencypytorch: CPU - 1 - ResNet-50pytorch: CPU - 1 - ResNet-152pytorch: CPU - 512 - ResNet-50pytorch: CPU - 512 - ResNet-152pytorch: CPU - 1 - Efficientnet_v2_lpytorch: CPU - 512 - Efficientnet_v2_ltensorflow: CPU - 1 - ResNet-50tensorflow: CPU - 512 - ResNet-50stress-ng: Hashstress-ng: MMAPstress-ng: NUMAstress-ng: Pipestress-ng: Pollstress-ng: Zlibstress-ng: Futexstress-ng: MEMFDstress-ng: Mutexstress-ng: Atomicstress-ng: Cryptostress-ng: Mallocstress-ng: Cloningstress-ng: Forkingstress-ng: Pthreadstress-ng: AVL Treestress-ng: IO_uringstress-ng: SENDFILEstress-ng: CPU Cachestress-ng: CPU Stressstress-ng: Semaphoresstress-ng: Matrix Mathstress-ng: Vector Mathstress-ng: AVX-512 VNNIstress-ng: Function Callstress-ng: x86_64 RdRandstress-ng: Floating Pointstress-ng: Matrix 3D Mathstress-ng: Memory Copyingstress-ng: Vector Shufflestress-ng: Mixed Schedulerstress-ng: Socket Activitystress-ng: Wide Vector Mathstress-ng: Context Switchingstress-ng: Fused Multiply-Addstress-ng: Vector Floating Pointstress-ng: Glibc C String Functionsstress-ng: Glibc Qsort Data Sortingstress-ng: System V Message Passingblender: BMW27 - CPU-Onlyblender: Barbershop - CPU-Onlyopenvino: Person Detection FP16 - CPUopenvino: Person Detection FP16 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Face Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16-INT8 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Person Vehicle Bike Detection FP16 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Handwritten English Recognition FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUrocksdb: Rand Readrocksdb: Update Randrocksdb: Read While Writingrocksdb: Read Rand Write Randspeedb: Rand Readspeedb: Update Randspeedb: 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.ggufllamafile: llava-v1.5-7b-q4 - CPUllamafile: mistral-7b-instruct-v0.2.Q8_0 - CPUllamafile: wizardcoder-python-34b-v1.0.Q6_K - CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9390.24353256278.614305107547.065334695.364492.5107.014072.834.614270.0876040136080720.439157.43586.82828.276168.09389.4788.5564.28436244640.2765421018.45123.6610.5721.088.786.542.332.04142.7642891632.604551.91691.4371470186.9613095189.4425361.452996767.15134.1246761181.37187.90492021.86339209442.2811924.3736691.2658661.313637.579895683.751475163.44855185.73460475.32301864524.10943747.151336243.8819310087.82158411.0072401267.3666221.4021674.0660720.85138701.98127531.88112726.265890325.7935325682.86441801423.41456920.91207322045.974661.0728034733.607.9870.761103.2143.48205.46233.3111840.174.0436011.875.314017.1811.941104.5643.3920939.799.099656.034.964622.4641.51158019.600.341103782203405246211988181594895111923593031166139636850162722521.1317.155.3621.3513.536.219389.41883656272.870281107670.840236694.664491.4106.654071.634.654266.4874679135174220.296153.72488.33627.598165.56187.6848.5904.38637297000.2685333918.74823.4710.3320.938.726.551.112.12143.2142841367.564685.26679.3461084286.2613211019.3125301.412243204.99834.7444546973.96187.72488061.57329314591.509532.6237276.2060556.773640.5312307832.192122557.99840635.00460897.89294597947.60944251.031332249.5619338319.36158205.8372323675.3966160.1420769.7760735.60138846.02127334.72124526.055893175.3242558881.72441769520.27456150.17204900453.754658.1528183544.827.9970.621105.1443.40205.21233.4111817.524.0536016.995.304017.9211.931101.2043.5320862.699.139648.434.964623.4841.50156517.910.341103280876411177170140881589032112843115531478632574249158508622.2717.173.5721.3713.476.219389.91588556280.203497107675.920990694.144491.7106.464074.134.684266.8889667135881719.970153.20887.25127.086161.38488.1208.6064.35435887550.2805787517.35419.238.2517.617.205.561.182.08144.0742856352.414627.00686.3761641558.0219277506.6125311.812410851.80840.5847405607.15183.27488178.36329981338.549747.4836748.3761798.903649.2318217916.692503374.61824396.72458555.64316273290.00928621.381333950.7019568145.68158272.4072276580.3866153.9921261.9860600.94138698.52133980.5266923.305886158.5942973852.02441959661.17456415.88201218048.984664.1120162835.908.0269.631099.9243.61205.45233.1011846.984.0435967.575.314008.8311.961111.8643.1020905.089.109678.994.954620.7941.52156137.760.341101418931406964167609461585552111743109031433232475216158093122.2315.703.4220.8513.616.219388.76468256268.964792107666.911599694.814490.4106.514073.634.684265.2876753135526119.905154.44387.51126.880156.23488.2088.6254.36337316480.2685611917.82119.268.3117.437.275.501.262.01145.7542902072.994270.84696.8266998085.9918907852.3125320.452337644.18658.6748020269.46182.38492474.65345731944.969729.1533018.3160918.243643.2515943494.882846032.14862417.29460093.34321848076.37947292.951331024.1219514410.29158008.6672193736.3265869.4821058.8160755.11138787.10134318.7166883.605884184.3041473670.27441825227.41456329.90196585232.624658.2528486279.677.9770.111099.2843.63205.32233.3911826.994.0535914.695.324007.1311.971115.4842.9620824.039.149658.444.964633.9741.40156828.340.341098560872432923150953551581099110290208332849033717802159405822.4516.495.2420.8113.526.23OpenBenchmarking.org

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: ReadLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan2K4K6K8K10KSE +/- 0.11, N = 3SE +/- 0.04, N = 3SE +/- 0.36, N = 3SE +/- 0.19, N = 39390.249389.429389.929388.76MIN: 9389.41 / MAX: 9390.99MIN: 9387.81 / MAX: 9390.39MIN: 9376.26 / MAX: 9391.12MIN: 9387.83 / MAX: 9389.931. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: WriteLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan12K24K36K48K60KSE +/- 0.29, N = 3SE +/- 0.98, N = 3SE +/- 0.28, N = 3SE +/- 0.22, N = 356278.6156272.8756280.2056268.96MIN: 56074.59 / MAX: 56343.25MIN: 56070.22 / MAX: 56338.93MIN: 56079.62 / MAX: 56345.05MIN: 56066.44 / MAX: 56333.31. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / WriteLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan20K40K60K80K100KSE +/- 39.89, N = 3SE +/- 145.98, N = 3SE +/- 36.94, N = 3SE +/- 77.76, N = 3107547.07107670.84107675.92107666.91MIN: 81047.89 / MAX: 111835.1MIN: 81031.82 / MAX: 111822.61MIN: 81045.61 / MAX: 111844.98MIN: 81021.88 / MAX: 1118251. (CC) gcc options: -O3 -lrt

LZ4 Compression

This test measures the time needed to compress/decompress a sample file (silesia archive) using LZ4 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 1 - Compression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan150300450600750SE +/- 0.25, N = 3SE +/- 0.52, N = 3SE +/- 0.90, N = 3SE +/- 1.35, N = 3695.36694.66694.14694.811. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 1 - Decompression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan10002000300040005000SE +/- 0.15, N = 3SE +/- 0.83, N = 3SE +/- 0.30, N = 3SE +/- 0.27, N = 34492.54491.44491.74490.41. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 3 - Compression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan20406080100SE +/- 0.09, N = 3SE +/- 0.03, N = 3SE +/- 0.21, N = 3SE +/- 0.10, N = 3107.01106.65106.46106.511. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 3 - Decompression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9001800270036004500SE +/- 0.78, N = 3SE +/- 0.61, N = 3SE +/- 0.49, N = 3SE +/- 0.29, N = 34072.84071.64074.14073.61. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 9 - Compression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan816243240SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 334.6134.6534.6834.681. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 9 - Decompression SpeedLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9001800270036004500SE +/- 0.07, N = 3SE +/- 0.95, N = 3SE +/- 0.41, N = 3SE +/- 0.90, N = 34270.04266.44266.84265.21. (CC) gcc options: -O3

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan200K400K600K800K1000KSE +/- 4619.10, N = 3SE +/- 2678.19, N = 3SE +/- 8666.07, N = 3SE +/- 9321.76, N = 38760408746798896678767531. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan300K600K900K1200K1500KSE +/- 3058.54, N = 3SE +/- 2795.86, N = 3SE +/- 5123.49, N = 3SE +/- 8322.15, N = 313608071351742135881713552611. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

Timed Apache Compilation

This test times how long it takes to build the Apache HTTPD web server. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Apache Compilation 2.4.41Time To CompileLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan510152025SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 320.4420.3019.9719.91

Timed Gem5 Compilation

This test times how long it takes to compile Gem5. Gem5 is a simulator for computer system architecture research. Gem5 is widely used for computer architecture research within the industry, academia, and more. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Gem5 Compilation 23.0.1Time To CompileLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan306090120150SE +/- 1.63, N = 4SE +/- 0.54, N = 3SE +/- 1.03, N = 3SE +/- 0.81, N = 3157.44153.72153.21154.44

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan20406080100SE +/- 0.26, N = 3SE +/- 1.10, N = 4SE +/- 0.19, N = 3SE +/- 0.78, N = 386.8388.3487.2587.51

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfigLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan714212835SE +/- 0.28, N = 6SE +/- 0.26, N = 7SE +/- 0.22, N = 9SE +/- 0.21, N = 1028.2827.6027.0926.88

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: allmodconfigLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4080120160200SE +/- 0.09, N = 3SE +/- 0.70, N = 3SE +/- 0.20, N = 3SE +/- 1.33, N = 3168.09165.56161.38156.23

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 19.8.1Time To CompileLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan20406080100SE +/- 0.67, N = 3SE +/- 0.11, N = 3SE +/- 0.51, N = 3SE +/- 0.43, N = 389.4887.6888.1288.21

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: 1BLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan246810SE +/- 0.013, N = 3SE +/- 0.011, N = 3SE +/- 0.009, N = 3SE +/- 0.014, N = 38.5568.5908.6068.625

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500MLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.98691.97382.96073.94764.9345SE +/- 0.047, N = 3SE +/- 0.006, N = 3SE +/- 0.010, N = 3SE +/- 0.010, N = 34.2844.3864.3544.363

PostgreSQL

This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read OnlyLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan800K1600K2400K3200K4000KSE +/- 35614.94, N = 12SE +/- 15124.24, N = 3SE +/- 68594.34, N = 12SE +/- 40627.80, N = 336244643729700358875537316481. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average LatencyLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.0630.1260.1890.2520.315SE +/- 0.003, N = 12SE +/- 0.001, N = 3SE +/- 0.006, N = 12SE +/- 0.003, N = 30.2760.2680.2800.2681. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read WriteLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan12K24K36K48K60KSE +/- 582.81, N = 3SE +/- 125.02, N = 3SE +/- 1379.59, N = 9SE +/- 395.45, N = 3542105333957875561191. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 16Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average LatencyLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan510152025SE +/- 0.20, N = 3SE +/- 0.04, N = 3SE +/- 0.40, N = 9SE +/- 0.13, N = 318.4518.7517.3517.821. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm

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-50Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan612182430SE +/- 0.34, N = 3SE +/- 0.23, N = 15SE +/- 0.20, N = 5SE +/- 0.16, N = 323.6623.4719.2319.26MIN: 13.25 / MAX: 25.04MIN: 9.72 / MAX: 25.75MIN: 11.52 / MAX: 20.48MIN: 11.04 / MAX: 20.23

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: ResNet-152Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3691215SE +/- 0.15, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 1110.5710.338.258.31MIN: 4.56 / MAX: 11.43MIN: 4.78 / MAX: 11.26MIN: 3.98 / MAX: 8.67MIN: 3.51 / MAX: 8.85

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-50Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan510152025SE +/- 0.14, N = 14SE +/- 0.06, N = 3SE +/- 0.12, N = 3SE +/- 0.19, N = 321.0820.9317.6117.43MIN: 11.86 / MAX: 22.44MIN: 13.14 / MAX: 22.13MIN: 11.32 / MAX: 18.41MIN: 9.78 / MAX: 18.32

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: ResNet-152Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan246810SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.07, N = 58.788.727.207.27MIN: 4.5 / MAX: 9.28MIN: 4.82 / MAX: 9.24MIN: 6.97 / MAX: 7.42MIN: 3.96 / MAX: 7.61

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan246810SE +/- 0.05, N = 10SE +/- 0.07, N = 12SE +/- 0.07, N = 3SE +/- 0.04, N = 36.546.555.565.50MIN: 3.15 / MAX: 7.04MIN: 3.01 / MAX: 7MIN: 3.52 / MAX: 5.71MIN: 2.96 / MAX: 5.62

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.1Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_lLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.52431.04861.57292.09722.6215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 92.331.111.181.26MIN: 1.64 / MAX: 2.77MIN: 0.49 / MAX: 2.12MIN: 0.47 / MAX: 2.46MIN: 0.49 / MAX: 2.38

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: ResNet-50Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.4770.9541.4311.9082.385SE +/- 0.03, N = 15SE +/- 0.01, N = 3SE +/- 0.02, N = 15SE +/- 0.02, N = 152.042.122.082.01

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 512 - Model: ResNet-50Linux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan306090120150SE +/- 0.50, N = 3SE +/- 0.15, N = 3SE +/- 0.48, N = 3SE +/- 0.29, N = 3142.76143.21144.07145.75

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: HashLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9M18M27M36M45MSE +/- 60053.64, N = 3SE +/- 49636.88, N = 3SE +/- 81559.25, N = 3SE +/- 37473.36, N = 342891632.6042841367.5642856352.4142902072.991. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: MMAPLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan10002000300040005000SE +/- 11.50, N = 3SE +/- 4.13, N = 3SE +/- 1.99, N = 3SE +/- 6.99, N = 34551.914685.264627.004270.841. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: NUMALinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan150300450600750SE +/- 5.42, N = 3SE +/- 1.38, N = 3SE +/- 3.34, N = 3SE +/- 3.92, N = 3691.43679.34686.37696.821. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: PipeLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan15M30M45M60M75MSE +/- 1232934.62, N = 12SE +/- 1231347.57, N = 15SE +/- 829037.53, N = 3SE +/- 812857.42, N = 1571470186.9661084286.2661641558.0266998085.991. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: PollLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4M8M12M16M20MSE +/- 2079.04, N = 3SE +/- 2956.12, N = 3SE +/- 3875.34, N = 3SE +/- 32342.64, N = 313095189.4413211019.3119277506.6118907852.311. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: ZlibLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan5K10K15K20K25KSE +/- 17.84, N = 3SE +/- 35.85, N = 3SE +/- 11.46, N = 3SE +/- 36.08, N = 325361.4525301.4125311.8125320.451. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: FutexLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan600K1200K1800K2400K3000KSE +/- 30946.87, N = 3SE +/- 35379.01, N = 15SE +/- 24021.26, N = 5SE +/- 28376.36, N = 152996767.152243204.992410851.802337644.181. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: MEMFDLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan2004006008001000SE +/- 1.98, N = 15SE +/- 2.46, N = 3SE +/- 0.33, N = 3SE +/- 5.11, N = 3134.12834.74840.58658.671. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: MutexLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan10M20M30M40M50MSE +/- 481173.64, N = 3SE +/- 560077.05, N = 3SE +/- 115602.55, N = 3SE +/- 286875.66, N = 346761181.3744546973.9647405607.1548020269.461. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AtomicLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4080120160200SE +/- 1.28, N = 3SE +/- 0.66, N = 3SE +/- 0.78, N = 3SE +/- 0.33, N = 3187.90187.72183.27182.381. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: CryptoLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan110K220K330K440K550KSE +/- 1454.92, N = 3SE +/- 2866.62, N = 3SE +/- 3698.19, N = 3SE +/- 710.36, N = 3492021.86488061.57488178.36492474.651. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: MallocLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan70M140M210M280M350MSE +/- 583008.10, N = 3SE +/- 113926.80, N = 3SE +/- 321332.84, N = 3SE +/- 2654426.38, N = 15339209442.28329314591.50329981338.54345731944.961. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: CloningLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3K6K9K12K15KSE +/- 102.00, N = 3SE +/- 1195.63, N = 12SE +/- 1944.59, N = 12SE +/- 1911.22, N = 1211924.379532.629747.489729.151. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: ForkingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan8K16K24K32K40KSE +/- 297.13, N = 13SE +/- 165.68, N = 3SE +/- 436.72, N = 15SE +/- 39.55, N = 336691.2637276.2036748.3733018.311. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: PthreadLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan13K26K39K52K65KSE +/- 246.97, N = 3SE +/- 806.76, N = 3SE +/- 486.25, N = 9SE +/- 224.09, N = 358661.3160556.7761798.9060918.241. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVL TreeLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan8001600240032004000SE +/- 21.80, N = 3SE +/- 12.35, N = 3SE +/- 9.76, N = 3SE +/- 9.73, N = 33637.573640.533649.233643.251. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: IO_uringLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4M8M12M16M20MSE +/- 2384203.44, N = 12SE +/- 1231639.25, N = 12SE +/- 245563.90, N = 3SE +/- 20817.51, N = 39895683.7512307832.1918217916.6915943494.881. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: SENDFILELinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan600K1200K1800K2400K3000KSE +/- 55032.25, N = 12SE +/- 24274.27, N = 4SE +/- 14114.72, N = 3SE +/- 16147.19, N = 31475163.442122557.992503374.612846032.141. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: CPU CacheLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan200K400K600K800K1000KSE +/- 957.96, N = 3SE +/- 1430.34, N = 3SE +/- 442.06, N = 3SE +/- 894.71, N = 3855185.73840635.00824396.72862417.291. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: CPU StressLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan100K200K300K400K500KSE +/- 1017.81, N = 3SE +/- 1508.14, N = 3SE +/- 1257.93, N = 3SE +/- 1639.51, N = 3460475.32460897.89458555.64460093.341. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: SemaphoresLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan70M140M210M280M350MSE +/- 698011.13, N = 3SE +/- 3210083.71, N = 3SE +/- 2323869.92, N = 3SE +/- 762255.36, N = 3301864524.10294597947.60316273290.00321848076.371. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Matrix MathLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan200K400K600K800K1000KSE +/- 2446.43, N = 3SE +/- 2651.09, N = 3SE +/- 2409.20, N = 3SE +/- 1530.59, N = 3943747.15944251.03928621.38947292.951. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector MathLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan300K600K900K1200K1500KSE +/- 554.48, N = 3SE +/- 990.33, N = 3SE +/- 635.24, N = 3SE +/- 1614.72, N = 31336243.881332249.561333950.701331024.121. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVX-512 VNNILinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4M8M12M16M20MSE +/- 34935.06, N = 3SE +/- 16761.27, N = 3SE +/- 16941.68, N = 3SE +/- 33202.31, N = 319310087.8219338319.3619568145.6819514410.291. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Function CallLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan30K60K90K120K150KSE +/- 110.30, N = 3SE +/- 166.05, N = 3SE +/- 70.27, N = 3SE +/- 190.36, N = 3158411.00158205.83158272.40158008.661. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: x86_64 RdRandLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan16M32M48M64M80MSE +/- 3744.59, N = 3SE +/- 2207.23, N = 3SE +/- 12777.82, N = 3SE +/- 16235.65, N = 372401267.3672323675.3972276580.3872193736.321. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Floating PointLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan14K28K42K56K70KSE +/- 56.56, N = 3SE +/- 86.63, N = 3SE +/- 67.12, N = 3SE +/- 42.50, N = 366221.4066160.1466153.9965869.481. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Matrix 3D MathLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan5K10K15K20K25KSE +/- 1665.45, N = 12SE +/- 1714.00, N = 12SE +/- 1654.57, N = 12SE +/- 1628.18, N = 1221674.0620769.7721261.9821058.811. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Memory CopyingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan13K26K39K52K65KSE +/- 52.89, N = 3SE +/- 99.29, N = 3SE +/- 50.44, N = 3SE +/- 82.66, N = 360720.8560735.6060600.9460755.111. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector ShuffleLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan30K60K90K120K150KSE +/- 149.16, N = 3SE +/- 59.67, N = 3SE +/- 242.92, N = 3SE +/- 149.78, N = 3138701.98138846.02138698.52138787.101. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Mixed SchedulerLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan30K60K90K120K150KSE +/- 1466.78, N = 4SE +/- 1093.17, N = 15SE +/- 1357.21, N = 15SE +/- 1003.39, N = 15127531.88127334.72133980.52134318.711. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Socket ActivityLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan30K60K90K120K150KSE +/- 940.39, N = 3SE +/- 208.73, N = 3SE +/- 414.89, N = 3SE +/- 310.90, N = 3112726.26124526.0566923.3066883.601. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Wide Vector MathLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1.3M2.6M3.9M5.2M6.5MSE +/- 1610.73, N = 3SE +/- 5891.75, N = 3SE +/- 5847.18, N = 3SE +/- 9854.66, N = 35890325.795893175.325886158.595884184.301. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Context SwitchingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9M18M27M36M45MSE +/- 735648.32, N = 15SE +/- 126646.22, N = 3SE +/- 58903.36, N = 3SE +/- 161811.56, N = 335325682.8642558881.7242973852.0241473670.271. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Fused Multiply-AddLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan90M180M270M360M450MSE +/- 119630.16, N = 3SE +/- 792348.84, N = 3SE +/- 1012815.75, N = 3SE +/- 775526.64, N = 3441801423.41441769520.27441959661.17441825227.411. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector Floating PointLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan100K200K300K400K500KSE +/- 485.94, N = 3SE +/- 411.82, N = 3SE +/- 491.76, N = 3SE +/- 350.71, N = 3456920.91456150.17456415.88456329.901. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Glibc C String FunctionsLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan40M80M120M160M200MSE +/- 1655443.15, N = 15SE +/- 1978439.64, N = 15SE +/- 1989236.60, N = 5SE +/- 789114.28, N = 3207322045.97204900453.75201218048.98196585232.621. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Glibc Qsort Data SortingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan10002000300040005000SE +/- 2.91, N = 3SE +/- 8.41, N = 3SE +/- 6.43, N = 3SE +/- 4.07, N = 34661.074658.154664.114658.251. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: System V Message PassingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan6M12M18M24M30MSE +/- 34165.06, N = 3SE +/- 32811.36, N = 3SE +/- 26751.58, N = 3SE +/- 17991.29, N = 328034733.6028183544.8220162835.9028486279.671. (CXX) g++ options: -lm -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lz

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0Blend File: BMW27 - Compute: CPU-OnlyLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan246810SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.07, N = 3SE +/- 0.01, N = 37.987.998.027.97

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.0Blend File: Barbershop - Compute: CPU-OnlyLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1632486480SE +/- 0.12, N = 3SE +/- 0.06, N = 3SE +/- 0.18, N = 3SE +/- 0.25, N = 370.7670.6269.6370.11

OpenVINO

This is a test of the Intel OpenVINO, a toolkit around neural networks, using its built-in benchmarking support and analyzing the throughput and latency for various models. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Person Detection FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan2004006008001000SE +/- 1.07, N = 3SE +/- 1.34, N = 3SE +/- 0.81, N = 3SE +/- 0.53, N = 31103.211105.141099.921099.281. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Person Detection FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1020304050SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 343.4843.4043.6143.63MIN: 37.8 / MAX: 109.72MIN: 40.58 / MAX: 93.21MIN: 39.44 / MAX: 93.31MIN: 40.69 / MAX: 93.91. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Face Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan50100150200250SE +/- 0.24, N = 3SE +/- 0.26, N = 3SE +/- 0.36, N = 3SE +/- 0.18, N = 3205.46205.21205.45205.321. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Face Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan50100150200250SE +/- 0.24, N = 3SE +/- 0.33, N = 3SE +/- 0.45, N = 3SE +/- 0.20, N = 3233.31233.41233.10233.39MIN: 207.82 / MAX: 247.74MIN: 206.8 / MAX: 293.22MIN: 207.01 / MAX: 346.48MIN: 207.86 / MAX: 270.461. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Vehicle Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3K6K9K12K15KSE +/- 14.65, N = 3SE +/- 16.20, N = 3SE +/- 8.61, N = 3SE +/- 2.19, N = 311840.1711817.5211846.9811826.991. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Vehicle Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.91131.82262.73393.64524.5565SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 34.044.054.044.05MIN: 3.59 / MAX: 22.24MIN: 3.62 / MAX: 19.96MIN: 3.59 / MAX: 25.59MIN: 3.64 / MAX: 25.831. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Face Detection Retail FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan8K16K24K32K40KSE +/- 27.67, N = 3SE +/- 40.16, N = 3SE +/- 5.33, N = 3SE +/- 25.38, N = 336011.8736016.9935967.5735914.691. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Face Detection Retail FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1.1972.3943.5914.7885.985SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 35.315.305.315.32MIN: 4.87 / MAX: 25.7MIN: 4.8 / MAX: 26.57MIN: 4.85 / MAX: 23.09MIN: 4.83 / MAX: 24.351. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Road Segmentation ADAS FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan9001800270036004500SE +/- 1.53, N = 3SE +/- 5.04, N = 3SE +/- 5.04, N = 3SE +/- 3.26, N = 34017.184017.924008.834007.131. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Road Segmentation ADAS FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 311.9411.9311.9611.97MIN: 10.72 / MAX: 42.19MIN: 10.66 / MAX: 47.97MIN: 11.05 / MAX: 41.55MIN: 11.11 / MAX: 39.941. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Machine Translation EN To DE FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan2004006008001000SE +/- 1.72, N = 3SE +/- 1.20, N = 3SE +/- 1.64, N = 3SE +/- 1.29, N = 31104.561101.201111.861115.481. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Machine Translation EN To DE FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1020304050SE +/- 0.07, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 3SE +/- 0.05, N = 343.3943.5343.1042.96MIN: 33.85 / MAX: 244.32MIN: 34.1 / MAX: 246.36MIN: 33.66 / MAX: 200.61MIN: 34.02 / MAX: 222.731. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan4K8K12K16K20KSE +/- 17.98, N = 3SE +/- 40.35, N = 3SE +/- 15.09, N = 3SE +/- 16.20, N = 320939.7920862.6920905.0820824.031. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Weld Porosity Detection FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3691215SE +/- 0.01, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 39.099.139.109.14MIN: 8.25 / MAX: 46.94MIN: 8.27 / MAX: 37.96MIN: 8.21 / MAX: 47.59MIN: 8.24 / MAX: 47.231. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan2K4K6K8K10KSE +/- 11.50, N = 3SE +/- 2.25, N = 3SE +/- 4.66, N = 3SE +/- 7.27, N = 39656.039648.439678.999658.441. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Person Vehicle Bike Detection FP16 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1.1162.2323.3484.4645.58SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 34.964.964.954.96MIN: 4.39 / MAX: 24.18MIN: 4.48 / MAX: 27.81MIN: 4.46 / MAX: 20.84MIN: 4.4 / MAX: 26.651. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Handwritten English Recognition FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan10002000300040005000SE +/- 24.48, N = 3SE +/- 10.46, N = 3SE +/- 7.21, N = 3SE +/- 6.46, N = 34622.464623.484620.794633.971. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Handwritten English Recognition FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan918273645SE +/- 0.22, N = 3SE +/- 0.09, N = 3SE +/- 0.06, N = 3SE +/- 0.06, N = 341.5141.5041.5241.40MIN: 37.79 / MAX: 60.99MIN: 38.11 / MAX: 59.33MIN: 37.94 / MAX: 61.47MIN: 38.14 / MAX: 58.291. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan30K60K90K120K150KSE +/- 1456.84, N = 3SE +/- 1446.63, N = 6SE +/- 294.88, N = 3SE +/- 1746.03, N = 3158019.60156517.91156137.76156828.341. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.2.devModel: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan0.07650.1530.22950.3060.3825SE +/- 0.00, N = 3SE +/- 0.00, N = 6SE +/- 0.00, N = 3SE +/- 0.00, N = 30.340.340.340.34MIN: 0.33 / MAX: 35.16MIN: 0.33 / MAX: 39.24MIN: 0.33 / MAX: 50.17MIN: 0.33 / MAX: 44.411. (CXX) g++ options: -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -pie

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Random ReadLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan200M400M600M800M1000MSE +/- 3926608.50, N = 3SE +/- 1094497.51, N = 3SE +/- 2253024.48, N = 3SE +/- 1037425.37, N = 311037822031103280876110141893110985608721. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Update RandomLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan90K180K270K360K450KSE +/- 1449.36, N = 3SE +/- 204.01, N = 3SE +/- 317.62, N = 3SE +/- 735.51, N = 34052464111774069644329231. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Read While WritingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan5M10M15M20M25MSE +/- 374386.25, N = 15SE +/- 531525.96, N = 15SE +/- 601741.45, N = 15SE +/- 428949.64, N = 15211988181701408816760946150953551. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Read Random Write RandomLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan300K600K900K1200K1500KSE +/- 12296.87, N = 3SE +/- 2394.21, N = 3SE +/- 1389.79, N = 3SE +/- 1372.82, N = 315948951589032158555215810991. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

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 ReadLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan200M400M600M800M1000MSE +/- 3522165.63, N = 3SE +/- 650677.38, N = 3SE +/- 2336203.11, N = 3SE +/- 14530585.11, N = 311192359301128431155111743109011029020831. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update RandomLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan70K140K210K280K350KSE +/- 2205.56, N = 3SE +/- 3481.85, N = 3SE +/- 968.55, N = 3SE +/- 263.27, N = 33116613147863143323284901. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While WritingLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan8M16M24M32M40MSE +/- 511361.34, N = 15SE +/- 840379.83, N = 15SE +/- 855245.74, N = 12SE +/- 1405761.06, N = 12396368503257424932475216337178021. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write RandomLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan300K600K900K1200K1500KSE +/- 8610.85, N = 3SE +/- 3477.16, N = 3SE +/- 12600.29, N = 3SE +/- 13303.55, N = 316272251585086158093115940581. (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.ggufLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan510152025SE +/- 0.50, N = 12SE +/- 0.11, N = 3SE +/- 0.09, N = 3SE +/- 0.20, N = 321.1322.2722.2322.451. (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.ggufLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan48121620SE +/- 0.10, N = 3SE +/- 0.05, N = 3SE +/- 0.58, N = 15SE +/- 0.47, N = 1217.1517.1715.7016.491. (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.ggufLinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan1.2062.4123.6184.8246.03SE +/- 0.01, N = 3SE +/- 0.22, N = 12SE +/- 0.12, N = 12SE +/- 0.01, N = 35.363.573.425.241. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -march=native -mtune=native -lopenblas

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: llava-v1.5-7b-q4 - Acceleration: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan510152025SE +/- 0.58, N = 15SE +/- 0.52, N = 15SE +/- 0.50, N = 15SE +/- 0.32, N = 1521.3521.3720.8520.81

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan3691215SE +/- 0.02, N = 3SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 313.5313.4713.6113.52

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPULinux 6.5Linux 6.6Linux 6.7Linux 6.8 26 Jan246810SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 36.216.216.216.23

106 Results Shown

CacheBench:
  Read
  Write
  Read / Modify / Write
LZ4 Compression:
  1 - Compression Speed
  1 - Decompression Speed
  3 - Compression Speed
  3 - Decompression Speed
  9 - Compression Speed
  9 - Decompression Speed
7-Zip Compression:
  Compression Rating
  Decompression Rating
Timed Apache Compilation
Timed Gem5 Compilation
Timed Godot Game Engine Compilation
Timed Linux Kernel Compilation:
  defconfig
  allmodconfig
Timed Node.js Compilation
Y-Cruncher:
  1B
  500M
PostgreSQL:
  100 - 1000 - Read Only
  100 - 1000 - Read Only - Average Latency
  100 - 1000 - Read Write
  100 - 1000 - Read Write - Average Latency
PyTorch:
  CPU - 1 - ResNet-50
  CPU - 1 - ResNet-152
  CPU - 512 - ResNet-50
  CPU - 512 - ResNet-152
  CPU - 1 - Efficientnet_v2_l
  CPU - 512 - Efficientnet_v2_l
TensorFlow:
  CPU - 1 - ResNet-50
  CPU - 512 - ResNet-50
Stress-NG:
  Hash
  MMAP
  NUMA
  Pipe
  Poll
  Zlib
  Futex
  MEMFD
  Mutex
  Atomic
  Crypto
  Malloc
  Cloning
  Forking
  Pthread
  AVL Tree
  IO_uring
  SENDFILE
  CPU Cache
  CPU Stress
  Semaphores
  Matrix Math
  Vector Math
  AVX-512 VNNI
  Function Call
  x86_64 RdRand
  Floating Point
  Matrix 3D Math
  Memory Copying
  Vector Shuffle
  Mixed Scheduler
  Socket Activity
  Wide Vector Math
  Context Switching
  Fused Multiply-Add
  Vector Floating Point
  Glibc C String Functions
  Glibc Qsort Data Sorting
  System V Message Passing
Blender:
  BMW27 - CPU-Only
  Barbershop - CPU-Only
OpenVINO:
  Person Detection FP16 - CPU:
    FPS
    ms
  Face Detection FP16-INT8 - CPU:
    FPS
    ms
  Vehicle Detection FP16-INT8 - CPU:
    FPS
    ms
  Face Detection Retail FP16-INT8 - CPU:
    FPS
    ms
  Road Segmentation ADAS FP16-INT8 - CPU:
    FPS
    ms
  Machine Translation EN To DE FP16 - CPU:
    FPS
    ms
  Weld Porosity Detection FP16-INT8 - CPU:
    FPS
    ms
  Person Vehicle Bike Detection FP16 - CPU:
    FPS
    ms
  Handwritten English Recognition FP16-INT8 - CPU:
    FPS
    ms
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU:
    FPS
    ms
RocksDB:
  Rand Read
  Update Rand
  Read While Writing
  Read Rand Write Rand
Speedb:
  Rand Read
  Update Rand
  Read While Writing
  Read Rand Write Rand
Llama.cpp:
  llama-2-7b.Q4_0.gguf
  llama-2-13b.Q4_0.gguf
  llama-2-70b-chat.Q5_0.gguf
Llamafile:
  llava-v1.5-7b-q4 - CPU
  mistral-7b-instruct-v0.2.Q8_0 - CPU
  wizardcoder-python-34b-v1.0.Q6_K - CPU