Ryzen 9 5950X Clang 12 vs. GCC 11 Benchmarks

GCC 11.1 versus LLVM Clang 12 on AMD Ryzen 9 5950X. 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 2105198-IB-11900KCOM08
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:

Audio Encoding 3 Tests
Bioinformatics 3 Tests
C/C++ Compiler Tests 12 Tests
CPU Massive 12 Tests
Creator Workloads 12 Tests
Encoding 6 Tests
HPC - High Performance Computing 6 Tests
Imaging 3 Tests
Machine Learning 2 Tests
MPI Benchmarks 2 Tests
Multi-Core 8 Tests
OpenMPI Tests 3 Tests
Renderers 2 Tests
Scientific Computing 4 Tests
Server CPU Tests 6 Tests
Single-Threaded 3 Tests
Video Encoding 3 Tests
Common Workstation Benchmarks 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
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Comparison
Transpose Comparison

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
GCC 11.1: -O2
May 18 2021
  4 Hours, 27 Minutes
GCC 11.1: -O3 -march=native
May 18 2021
  4 Hours, 36 Minutes
GCC 11.1: -O3 -march=native -flto
May 18 2021
  1 Hour, 47 Minutes
Clang 12: -O2
May 17 2021
  1 Hour, 43 Minutes
Clang 12: -O3 -march=native
May 17 2021
  1 Hour, 42 Minutes
Clang 12: -O3 -march=native -flto
May 18 2021
  1 Hour, 32 Minutes
Invert Hiding All Results Option
  2 Hours, 38 Minutes

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):


Ryzen 9 5950X Clang 12 vs. GCC 11 BenchmarksOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads)ASUS ROG CROSSHAIR VIII HERO (WI-FI) (3302 BIOS)AMD Starship/Matisse32GB500GB Western Digital WDS500G3X0C-00SJG0AMD Radeon RX 5600 OEM/5600 XT / 5700/5700 8GB (2100/875MHz)AMD Navi 10 HDMI AudioASUS MG28URealtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200Fedora 345.11.20-300.fc34.x86_64 (x86_64)GNOME Shell 40.1X Server + Wayland4.6 Mesa 21.0.3 (LLVM 12.0.0)GCC 11.1.1 20210428 + Clang 12.0.0Clang 12.0.0btrfs3840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilersFile-SystemScreen ResolutionRyzen 9 5950X Clang 12 Vs. GCC 11 Benchmarks PerformanceSystem Logs- Transparent Huge Pages: madvise- GCC 11.1: -O2: CXXFLAGS=-O2 CFLAGS=-O2- GCC 11.1: -O3 -march=native: CXXFLAGS="-O3 -march=native" CFLAGS="-O3 -march=native"- GCC 11.1: -O3 -march=native -flto: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto" - Clang 12: -O2: CXXFLAGS=-O2 CFLAGS=-O2- Clang 12: -O3 -march=native: CXXFLAGS="-O3 -march=native" CFLAGS="-O3 -march=native"- Clang 12: -O3 -march=native -flto: CXXFLAGS="-O3 -march=native -flto" CFLAGS="-O3 -march=native -flto" - GCC 11.1: -O2, GCC 11.1: -O3 -march=native, GCC 11.1: -O3 -march=native -flto: --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,lto --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=i686 --with-gcc-major-version-only --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa201009- SELinux + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

mrbayes: Primate Phylogeny Analysishmmer: Pfam Database Searchlammps: Rhodopsin Proteinwebp: Quality 100, Losslesswebp: Quality 100, Highest Compressionwebp: Quality 100, Lossless, Highest Compressiongraphics-magick: Rotategraphics-magick: Sharpengraphics-magick: Enhancedgraphics-magick: Resizingsvt-hevc: 7 - Bosphorus 1080psvt-hevc: 10 - Bosphorus 1080psvt-vp9: VMAF Optimized - Bosphorus 1080psvt-vp9: PSNR/SSIM Optimized - Bosphorus 1080psvt-vp9: Visual Quality Optimized - Bosphorus 1080px265: Bosphorus 4Kcoremark: CoreMark Size 666 - Iterations Per Secondhimeno: Poisson Pressure Solverpjsip: INVITEpjsip: OPTIONS, Statefulpjsip: OPTIONS, Statelessc-ray: Total Time - 4K, 16 Rays Per Pixelaobench: 2048 x 2048 - Total Timeencode-flac: WAV To FLACencode-mp3: WAV To MP3encode-opus: WAV To Opus Encodeliquid-dsp: 8 - 256 - 57liquid-dsp: 16 - 256 - 57tjbench: Decompression Throughputastcenc: Mediumastcenc: Thoroughastcenc: Exhaustivesqlite-speedtest: Timed Time - Size 1,000ncnn: CPU - mobilenetncnn: CPU-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU - mnasnetncnn: CPU - efficientnet-b0ncnn: CPU - googlenetncnn: CPU - vgg16ncnn: CPU - resnet18ncnn: CPU - alexnetncnn: CPU - resnet50ncnn: CPU - squeezenet_ssdncnn: CPU - regnety_400mtnn: CPU - MobileNet v2tnn: CPU - SqueezeNet v1.1ncnn: CPU - shufflenet-v2ncnn: CPU - blazefacencnn: CPU - yolov4-tinyGCC 11.1Clang 12 -O2 -O3 -march=native -O3 -march=native -flto -O2 -O3 -march=native -O3 -march=native -flto96.78599.72212.85513.1815.35527.4249852264221803217.32365.79230.53235.27222.6326.24830811.5321165120.9402624815786023059960.93131.7325.8746.8026.5155948500001041200000277.4968024.62967.684157.144547.40412.954.544.153.975.4113.1556.8214.5011.1626.1814.6117.24219.130208.7804.331.8521.0692.97397.74412.74913.5505.20028.32710363704492120221.71369.54231.16235.59223.1125.91808580.8626505314.7339044671798222175925.40525.4796.2375.5035.3876115600001085000000268.1134564.62777.571856.641146.91712.704.433.843.955.3712.8557.6114.3310.9925.3313.9317.18227.768214.2774.391.8021.0093.33797.71412.81313.2545.18427.9499643824492085227.04368.55230.43234.70222.3326.17849671.9594425445.6352214616791622258125.48226.3126.1955.4035.4756193933331093266667270.9091504.69957.529956.190046.47913.744.463.863.925.3613.2657.8414.8011.1325.8414.5018.53215.341202.5726.012.5523.9493.52996.09213.08713.6474.77829.1199212064111693225.15368.55231.35235.64225.5727.77712758.3578264877.3064104575794222110749.27533.7787.6876.5655.8745766600001056400000267.5530683.33679.395052.204847.79211.583.563.363.264.5211.8857.2614.0511.0922.9912.6112.93271.028208.5553.811.5421.9790.78595.21413.29914.1054.70328.6409812354521767229.27375.95234.71239.88227.7327.75722694.0062755192.1900424708795822257245.01030.2345.6666.0985.5085760500001061900000272.5135593.31799.354951.823047.63011.463.573.073.264.5111.7756.9313.9011.1623.2512.3512.59355.925215.7813.741.5221.7392.40593.31113.34213.6364.72227.4009612374511762236.05375.08232.66238.13227.6628.47714933.8589835289.91757144.74229.7305.7565.8105.5585878000001073466667266.2298293.37289.443251.897448.58411.053.422.933.084.3011.6256.1113.8611.2522.9012.4312.14261.836207.9253.691.4621.58OpenBenchmarking.org

Timed MrBayes Analysis

This test performs a bayesian analysis of a set of primate genome sequences in order to estimate their phylogeny. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MrBayes Analysis 3.2.7Primate Phylogeny Analysis-O2-O3 -march=native-O3 -march=native -flto20406080100SE +/- 0.21, N = 3SE +/- 0.18, N = 3SE +/- 0.19, N = 3SE +/- 0.13, N = 3SE +/- 0.29, N = 3SE +/- 0.13, N = 393.5390.7992.4196.7992.9793.341. (CC) gcc options: -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -msha -maes -mavx -mfma -mavx2 -mrdrnd -mbmi -mbmi2 -madx -mabm -O3 -std=c99 -pedantic -lm
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MrBayes Analysis 3.2.7Primate Phylogeny Analysis-O2-O3 -march=native-O3 -march=native -flto20406080100Min: 93.16 / Avg: 93.53 / Max: 93.87Min: 90.46 / Avg: 90.79 / Max: 91.09Min: 92.06 / Avg: 92.4 / Max: 92.72Min: 96.53 / Avg: 96.79 / Max: 96.92Min: 92.46 / Avg: 92.97 / Max: 93.46Min: 93.11 / Avg: 93.34 / Max: 93.561. (CC) gcc options: -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -msha -maes -mavx -mfma -mavx2 -mrdrnd -mbmi -mbmi2 -madx -mabm -O3 -std=c99 -pedantic -lm

Timed HMMer Search

This test searches through the Pfam database of profile hidden markov models. The search finds the domain structure of Drosophila Sevenless protein. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 3.3.2Pfam Database Search-O2-O3 -march=native-O3 -march=native -flto20406080100SE +/- 0.08, N = 3SE +/- 0.11, N = 3SE +/- 0.08, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 396.0995.2193.3199.7297.7497.711. (CC) gcc options: -pthread -lhmmer -leasel -lm -lmpi
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterTimed HMMer Search 3.3.2Pfam Database Search-O2-O3 -march=native-O3 -march=native -flto20406080100Min: 95.93 / Avg: 96.09 / Max: 96.2Min: 95.06 / Avg: 95.21 / Max: 95.42Min: 93.16 / Avg: 93.31 / Max: 93.44Min: 99.64 / Avg: 99.72 / Max: 99.78Min: 97.68 / Avg: 97.74 / Max: 97.83Min: 97.56 / Avg: 97.71 / Max: 97.81. (CC) gcc options: -pthread -lhmmer -leasel -lm -lmpi

LAMMPS Molecular Dynamics Simulator

LAMMPS is a classical molecular dynamics code, and an acronym for Large-scale Atomic/Molecular Massively Parallel Simulator. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: Rhodopsin Protein-O2-O3 -march=native-O3 -march=native -flto3691215SE +/- 0.08, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.11, N = 8SE +/- 0.14, N = 4SE +/- 0.09, N = 1513.0913.3013.3412.8612.7512.811. (CXX) g++ options: -O2 -pthread -lm
Clang 12GCC 11.1OpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: Rhodopsin Protein-O2-O3 -march=native-O3 -march=native -flto48121620Min: 12.95 / Avg: 13.09 / Max: 13.23Min: 13.26 / Avg: 13.3 / Max: 13.33Min: 13.33 / Avg: 13.34 / Max: 13.35Min: 12.24 / Avg: 12.86 / Max: 13.11Min: 12.32 / Avg: 12.75 / Max: 12.91Min: 12.17 / Avg: 12.81 / Max: 13.341. (CXX) g++ options: -O2 -pthread -lm

WebP Image Encode

This is a test of Google's libwebp with the cwebp image encode utility and using a sample 6000x4000 pixel JPEG image as the input. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless-O2-O3 -march=native-O3 -march=native -flto48121620SE +/- 0.19, N = 3SE +/- 0.05, N = 3SE +/- 0.15, N = 4SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.17, N = 313.6514.1113.6413.1813.5513.251. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg
Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless-O2-O3 -march=native-O3 -march=native -flto48121620Min: 13.35 / Avg: 13.65 / Max: 13.99Min: 14.06 / Avg: 14.11 / Max: 14.2Min: 13.27 / Avg: 13.64 / Max: 13.98Min: 13.11 / Avg: 13.18 / Max: 13.22Min: 13.44 / Avg: 13.55 / Max: 13.61Min: 12.94 / Avg: 13.25 / Max: 13.51. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg

Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest Compression-O2-O3 -march=native-O3 -march=native -flto1.20492.40983.61474.81966.0245SE +/- 0.059, N = 4SE +/- 0.026, N = 3SE +/- 0.027, N = 3SE +/- 0.021, N = 3SE +/- 0.023, N = 3SE +/- 0.036, N = 34.7784.7034.7225.3555.2005.1841. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg
Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Highest Compression-O2-O3 -march=native-O3 -march=native -flto246810Min: 4.7 / Avg: 4.78 / Max: 4.95Min: 4.67 / Avg: 4.7 / Max: 4.75Min: 4.67 / Avg: 4.72 / Max: 4.76Min: 5.32 / Avg: 5.35 / Max: 5.39Min: 5.18 / Avg: 5.2 / Max: 5.25Min: 5.15 / Avg: 5.18 / Max: 5.261. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg

Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest Compression-O2-O3 -march=native-O3 -march=native -flto714212835SE +/- 0.13, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 329.1228.6427.4027.4228.3327.951. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg
Clang 12GCC 11.1OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100, Lossless, Highest Compression-O2-O3 -march=native-O3 -march=native -flto612182430Min: 28.9 / Avg: 29.12 / Max: 29.34Min: 28.54 / Avg: 28.64 / Max: 28.71Min: 27.37 / Avg: 27.4 / Max: 27.44Min: 27.36 / Avg: 27.42 / Max: 27.46Min: 28.3 / Avg: 28.33 / Max: 28.35Min: 27.81 / Avg: 27.95 / Max: 28.031. (CC) gcc options: -fvisibility=hidden -pthread -lm -ljpeg

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample 6000x4000 pixel JPEG image. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Rotate-O2-O3 -march=native-O3 -march=native -flto2004006008001000SE +/- 5.55, N = 3SE +/- 0.58, N = 3SE +/- 4.37, N = 3SE +/- 11.70, N = 3SE +/- 5.57, N = 3SE +/- 3.48, N = 392198196198510369641. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Rotate-O2-O3 -march=native-O3 -march=native -flto2004006008001000Min: 911 / Avg: 921.33 / Max: 930Min: 980 / Avg: 981 / Max: 982Min: 956 / Avg: 961.33 / Max: 970Min: 969 / Avg: 985.33 / Max: 1008Min: 1029 / Avg: 1036 / Max: 1047Min: 958 / Avg: 963.67 / Max: 9701. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Sharpen-O2-O3 -march=native-O3 -march=native -flto80160240320400SE +/- 0.58, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 0.33, N = 3SE +/- 1.20, N = 3SE +/- 0.88, N = 32062352372263703821. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Sharpen-O2-O3 -march=native-O3 -march=native -flto70140210280350Min: 205 / Avg: 206 / Max: 207Min: 235 / Avg: 235.33 / Max: 236Min: 236 / Avg: 236.67 / Max: 238Min: 226 / Avg: 226.33 / Max: 227Min: 368 / Avg: 369.67 / Max: 372Min: 381 / Avg: 382.33 / Max: 3841. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Enhanced-O2-O3 -march=native-O3 -march=native -flto100200300400500SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 34114524514224494491. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Enhanced-O2-O3 -march=native-O3 -march=native -flto80160240320400Min: 411 / Avg: 411.33 / Max: 412Min: 452 / Avg: 452.33 / Max: 453Min: 422 / Avg: 422.33 / Max: 423Min: 449 / Avg: 449.33 / Max: 4501. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Resizing-O2-O3 -march=native-O3 -march=native -flto5001000150020002500SE +/- 4.48, N = 3SE +/- 3.61, N = 3SE +/- 2.85, N = 3SE +/- 0.88, N = 3SE +/- 1.53, N = 3SE +/- 3.71, N = 31693176717621803212020851. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Resizing-O2-O3 -march=native-O3 -march=native -flto400800120016002000Min: 1687 / Avg: 1693.33 / Max: 1702Min: 1760 / Avg: 1767 / Max: 1772Min: 1759 / Avg: 1762.33 / Max: 1768Min: 1802 / Avg: 1803.33 / Max: 1805Min: 2117 / Avg: 2120 / Max: 2122Min: 2080 / Avg: 2084.67 / Max: 20921. (CC) gcc options: -fopenmp -pthread -ljpeg -lz -lm -lpthread

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto50100150200250SE +/- 0.95, N = 3SE +/- 0.32, N = 3SE +/- 1.15, N = 3SE +/- 0.49, N = 3SE +/- 0.83, N = 3SE +/- 0.03, N = 3225.15229.27236.05217.32221.71227.041. (CC) gcc options: -O2 -fPIE -fPIC -O3 -pie -rdynamic -lpthread -lrt
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto4080120160200Min: 223.3 / Avg: 225.15 / Max: 226.42Min: 228.83 / Avg: 229.27 / Max: 229.89Min: 233.74 / Avg: 236.05 / Max: 237.25Min: 216.61 / Avg: 217.32 / Max: 218.26Min: 220.51 / Avg: 221.71 / Max: 223.3Min: 227.01 / Avg: 227.04 / Max: 227.11. (CC) gcc options: -O2 -fPIE -fPIC -O3 -pie -rdynamic -lpthread -lrt

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto80160240320400SE +/- 0.23, N = 3SE +/- 1.66, N = 3SE +/- 0.90, N = 3SE +/- 1.24, N = 3SE +/- 1.07, N = 3SE +/- 0.57, N = 3368.55375.95375.08365.79369.54368.551. (CC) gcc options: -O2 -fPIE -fPIC -O3 -pie -rdynamic -lpthread -lrt
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto70140210280350Min: 368.32 / Avg: 368.55 / Max: 369Min: 374.06 / Avg: 375.95 / Max: 379.27Min: 373.37 / Avg: 375.08 / Max: 376.41Min: 363.86 / Avg: 365.79 / Max: 368.1Min: 367.87 / Avg: 369.54 / Max: 371.52Min: 367.42 / Avg: 368.55 / Max: 369.231. (CC) gcc options: -O2 -fPIE -fPIC -O3 -pie -rdynamic -lpthread -lrt

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample YUV input video file. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto50100150200250SE +/- 0.33, N = 3SE +/- 0.32, N = 3SE +/- 0.70, N = 3SE +/- 0.09, N = 3SE +/- 0.46, N = 3SE +/- 0.18, N = 3231.35234.71232.66230.53231.16230.431. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto4080120160200Min: 230.7 / Avg: 231.35 / Max: 231.74Min: 234.14 / Avg: 234.71 / Max: 235.25Min: 231.27 / Avg: 232.66 / Max: 233.41Min: 230.38 / Avg: 230.53 / Max: 230.69Min: 230.32 / Avg: 231.16 / Max: 231.91Min: 230.09 / Avg: 230.43 / Max: 230.681. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto50100150200250SE +/- 0.18, N = 3SE +/- 0.40, N = 3SE +/- 0.48, N = 3SE +/- 0.37, N = 3SE +/- 0.23, N = 3SE +/- 0.33, N = 3235.64239.88238.13235.27235.59234.701. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto4080120160200Min: 235.41 / Avg: 235.64 / Max: 235.99Min: 239.09 / Avg: 239.88 / Max: 240.36Min: 237.28 / Avg: 238.13 / Max: 238.95Min: 234.53 / Avg: 235.27 / Max: 235.66Min: 235.36 / Avg: 235.59 / Max: 236.05Min: 234.33 / Avg: 234.7 / Max: 235.351. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto50100150200250SE +/- 1.28, N = 3SE +/- 0.75, N = 3SE +/- 0.89, N = 3SE +/- 0.82, N = 3SE +/- 0.87, N = 3SE +/- 0.97, N = 3225.57227.73227.66222.63223.11222.331. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080p-O2-O3 -march=native-O3 -march=native -flto4080120160200Min: 223.09 / Avg: 225.57 / Max: 227.38Min: 226.84 / Avg: 227.73 / Max: 229.22Min: 226.33 / Avg: 227.66 / Max: 229.36Min: 221.55 / Avg: 222.63 / Max: 224.25Min: 221.38 / Avg: 223.11 / Max: 224.01Min: 220.76 / Avg: 222.33 / Max: 224.091. (CC) gcc options: -O3 -fcommon -O2 -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4K-O2-O3 -march=native-O3 -march=native -flto714212835SE +/- 0.12, N = 3SE +/- 0.04, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 327.7727.7528.4726.2425.9126.171. (CXX) g++ options: -O2 -rdynamic -lpthread -lrt -ldl
Clang 12GCC 11.1OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4K-O2-O3 -march=native-O3 -march=native -flto612182430Min: 27.52 / Avg: 27.77 / Max: 27.91Min: 27.68 / Avg: 27.75 / Max: 27.8Min: 28.36 / Avg: 28.47 / Max: 28.67Min: 26.14 / Avg: 26.24 / Max: 26.34Min: 25.84 / Avg: 25.91 / Max: 25.96Min: 25.99 / Avg: 26.17 / Max: 26.261. (CXX) g++ options: -O2 -rdynamic -lpthread -lrt -ldl

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per Second-O2-O3 -march=native-O3 -march=native -flto200K400K600K800K1000KSE +/- 789.22, N = 3SE +/- 845.83, N = 3SE +/- 581.76, N = 3SE +/- 989.68, N = 3SE +/- 2230.96, N = 3SE +/- 836.68, N = 3712758.36722694.01714933.86830811.53808580.86849671.961. (CC) gcc options: -O2 -lrt" -lrt
Clang 12GCC 11.1OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per Second-O2-O3 -march=native-O3 -march=native -flto150K300K450K600K750KMin: 711532.76 / Avg: 712758.36 / Max: 714232.57Min: 721451.92 / Avg: 722694.01 / Max: 724309.64Min: 714338.86 / Avg: 714933.86 / Max: 716097.27Min: 829015.54 / Avg: 830811.53 / Max: 832430.09Min: 805707.09 / Avg: 808580.86 / Max: 812973.71Min: 848019.08 / Avg: 849671.96 / Max: 850724.451. (CC) gcc options: -O2 -lrt" -lrt

Himeno Benchmark

The Himeno benchmark is a linear solver of pressure Poisson using a point-Jacobi method. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure Solver-O2-O3 -march=native-O3 -march=native -flto12002400360048006000SE +/- 62.44, N = 15SE +/- 66.92, N = 15SE +/- 96.57, N = 13SE +/- 83.12, N = 15SE +/- 90.91, N = 15SE +/- 68.97, N = 154877.315192.195289.925120.945314.735445.641. (CC) gcc options: -O3 -mavx2
Clang 12GCC 11.1OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure Solver-O2-O3 -march=native-O3 -march=native -flto9001800270036004500Min: 4561.71 / Avg: 4877.31 / Max: 5218.53Min: 4803.14 / Avg: 5192.19 / Max: 5645.17Min: 4873.13 / Avg: 5289.92 / Max: 5793.99Min: 4712.52 / Avg: 5120.94 / Max: 5619.83Min: 4768.11 / Avg: 5314.73 / Max: 5885.31Min: 4944.37 / Avg: 5445.64 / Max: 5835.51. (CC) gcc options: -O3 -mavx2

PJSIP

PJSIP is a free and open source multimedia communication library written in C language implementing standard based protocols such as SIP, SDP, RTP, STUN, TURN, and ICE. It combines signaling protocol (SIP) with rich multimedia framework and NAT traversal functionality into high level API that is portable and suitable for almost any type of systems ranging from desktops, embedded systems, to mobile handsets. This test profile is making use of pjsip-perf with both the client/server on teh system. More details on the PJSIP benchmark at https://www.pjsip.org/high-performance-sip.htm Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: INVITE-O2-O3 -march=native-O3 -march=native -flto10002000300040005000SE +/- 42.78, N = 3SE +/- 66.36, N = 3SE +/- 16.33, N = 3SE +/- 32.67, N = 15SE +/- 65.77, N = 3457547084815467146161. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: INVITE-O2-O3 -march=native-O3 -march=native -flto8001600240032004000Min: 4509 / Avg: 4574.67 / Max: 4655Min: 4575 / Avg: 4707.67 / Max: 4777Min: 4784 / Avg: 4815.33 / Max: 4839Min: 4419 / Avg: 4671 / Max: 4908Min: 4518 / Avg: 4616 / Max: 47411. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, Stateful-O2-O3 -march=native-O3 -march=native -flto2K4K6K8K10KSE +/- 27.78, N = 3SE +/- 47.54, N = 3SE +/- 72.34, N = 3SE +/- 21.79, N = 3SE +/- 56.75, N = 3794279587860798279161. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, Stateful-O2-O3 -march=native-O3 -march=native -flto14002800420056007000Min: 7896 / Avg: 7942 / Max: 7992Min: 7880 / Avg: 7957.67 / Max: 8044Min: 7717 / Avg: 7860.33 / Max: 7949Min: 7956 / Avg: 7981.67 / Max: 8025Min: 7814 / Avg: 7916.33 / Max: 80101. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, Stateless-O2-O3 -march=native-O3 -march=native -flto50K100K150K200K250KSE +/- 1827.48, N = 3SE +/- 962.04, N = 3SE +/- 586.86, N = 3SE +/- 1749.48, N = 3SE +/- 705.50, N = 32211072225722305992217592225811. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgResponses Per Second, More Is BetterPJSIP 2.11Method: OPTIONS, Stateless-O2-O3 -march=native-O3 -march=native -flto40K80K120K160K200KMin: 218415 / Avg: 221107 / Max: 224594Min: 220779 / Avg: 222572 / Max: 224073Min: 229519 / Avg: 230599 / Max: 231537Min: 218922 / Avg: 221758.67 / Max: 224951Min: 221174 / Avg: 222581.33 / Max: 2233731. (CC) gcc options: -lstdc++ -lssl -lcrypto -lm -lrt -lpthread

C-Ray

This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per Pixel-O2-O3 -march=native-O3 -march=native -flto1428425670SE +/- 0.16, N = 3SE +/- 0.10, N = 3SE +/- 0.17, N = 3SE +/- 0.18, N = 3SE +/- 0.08, N = 3SE +/- 0.08, N = 349.2845.0144.7460.9325.4125.481. (CC) gcc options: -lm -lpthread -O3
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per Pixel-O2-O3 -march=native-O3 -march=native -flto1224364860Min: 48.97 / Avg: 49.27 / Max: 49.51Min: 44.82 / Avg: 45.01 / Max: 45.16Min: 44.45 / Avg: 44.74 / Max: 45.03Min: 60.58 / Avg: 60.93 / Max: 61.17Min: 25.26 / Avg: 25.41 / Max: 25.53Min: 25.33 / Avg: 25.48 / Max: 25.61. (CC) gcc options: -lm -lpthread -O3

AOBench

AOBench is a lightweight ambient occlusion renderer, written in C. The test profile is using a size of 2048 x 2048. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total Time-O2-O3 -march=native-O3 -march=native -flto816243240SE +/- 0.04, N = 3SE +/- 0.43, N = 3SE +/- 0.09, N = 3SE +/- 0.14, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 333.7830.2329.7331.7325.4826.311. (CC) gcc options: -lm -O3
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total Time-O2-O3 -march=native-O3 -march=native -flto714212835Min: 33.72 / Avg: 33.78 / Max: 33.85Min: 29.42 / Avg: 30.23 / Max: 30.88Min: 29.62 / Avg: 29.73 / Max: 29.91Min: 31.54 / Avg: 31.73 / Max: 32Min: 25.45 / Avg: 25.48 / Max: 25.53Min: 26.29 / Avg: 26.31 / Max: 26.331. (CC) gcc options: -lm -O3

FLAC Audio Encoding

This test times how long it takes to encode a sample WAV file to FLAC format five times. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.2WAV To FLAC-O2-O3 -march=native-O3 -march=native -flto246810SE +/- 0.024, N = 5SE +/- 0.005, N = 5SE +/- 0.030, N = 5SE +/- 0.045, N = 5SE +/- 0.020, N = 5SE +/- 0.036, N = 57.6875.6665.7565.8746.2376.1951. (CXX) g++ options: -fvisibility=hidden -logg -lm
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.2WAV To FLAC-O2-O3 -march=native-O3 -march=native -flto3691215Min: 7.61 / Avg: 7.69 / Max: 7.74Min: 5.65 / Avg: 5.67 / Max: 5.68Min: 5.69 / Avg: 5.76 / Max: 5.83Min: 5.79 / Avg: 5.87 / Max: 5.99Min: 6.16 / Avg: 6.24 / Max: 6.27Min: 6.09 / Avg: 6.2 / Max: 6.281. (CXX) g++ options: -fvisibility=hidden -logg -lm

LAME MP3 Encoding

LAME is an MP3 encoder licensed under the LGPL. This test measures the time required to encode a WAV file to MP3 format. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3-O2-O3 -march=native-O3 -march=native -flto246810SE +/- 0.068, N = 4SE +/- 0.045, N = 3SE +/- 0.056, N = 3SE +/- 0.084, N = 3SE +/- 0.071, N = 3SE +/- 0.010, N = 36.5656.0985.8106.8025.5035.4031. (CC) gcc options: -O3 -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr -pipe -lm
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3-O2-O3 -march=native-O3 -march=native -flto3691215Min: 6.38 / Avg: 6.57 / Max: 6.67Min: 6.05 / Avg: 6.1 / Max: 6.19Min: 5.7 / Avg: 5.81 / Max: 5.89Min: 6.64 / Avg: 6.8 / Max: 6.93Min: 5.37 / Avg: 5.5 / Max: 5.61Min: 5.39 / Avg: 5.4 / Max: 5.421. (CC) gcc options: -O3 -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr -pipe -lm

Opus Codec Encoding

Opus is an open audio codec. Opus is a lossy audio compression format designed primarily for interactive real-time applications over the Internet. This test uses Opus-Tools and measures the time required to encode a WAV file to Opus. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterOpus Codec Encoding 1.3.1WAV To Opus Encode-O2-O3 -march=native-O3 -march=native -flto246810SE +/- 0.052, N = 5SE +/- 0.045, N = 5SE +/- 0.016, N = 5SE +/- 0.035, N = 5SE +/- 0.048, N = 5SE +/- 0.023, N = 55.8745.5085.5586.5155.3875.4751. (CXX) g++ options: -fvisibility=hidden -logg -lm
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterOpus Codec Encoding 1.3.1WAV To Opus Encode-O2-O3 -march=native-O3 -march=native -flto3691215Min: 5.68 / Avg: 5.87 / Max: 5.98Min: 5.35 / Avg: 5.51 / Max: 5.6Min: 5.52 / Avg: 5.56 / Max: 5.6Min: 6.43 / Avg: 6.52 / Max: 6.6Min: 5.25 / Avg: 5.39 / Max: 5.48Min: 5.42 / Avg: 5.48 / Max: 5.531. (CXX) g++ options: -fvisibility=hidden -logg -lm

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 8 - Buffer Length: 256 - Filter Length: 57-O2-O3 -march=native-O3 -march=native -flto130M260M390M520M650MSE +/- 1824399.44, N = 3SE +/- 4072558.57, N = 3SE +/- 1855020.22, N = 3SE +/- 185202.59, N = 3SE +/- 4455270.29, N = 3SE +/- 1343755.10, N = 35766600005760500005878000005948500006115600006193933331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Clang 12GCC 11.1OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 8 - Buffer Length: 256 - Filter Length: 57-O2-O3 -march=native-O3 -march=native -flto110M220M330M440M550MMin: 574300000 / Avg: 576660000 / Max: 580250000Min: 567910000 / Avg: 576050000 / Max: 580370000Min: 585930000 / Avg: 587800000 / Max: 591510000Min: 594500000 / Avg: 594850000 / Max: 595130000Min: 603250000 / Avg: 611560000 / Max: 618500000Min: 617130000 / Avg: 619393333.33 / Max: 6217800001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Clang 12GCC 11.1OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 16 - Buffer Length: 256 - Filter Length: 57-O2-O3 -march=native-O3 -march=native -flto200M400M600M800M1000MSE +/- 2690724.81, N = 3SE +/- 493288.29, N = 3SE +/- 3219903.38, N = 3SE +/- 4650806.38, N = 3SE +/- 1252996.41, N = 3SE +/- 783865.07, N = 31056400000106190000010734666671041200000108500000010932666671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Clang 12GCC 11.1OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 2021.01.31Threads: 16 - Buffer Length: 256 - Filter Length: 57-O2-O3 -march=native-O3 -march=native -flto200M400M600M800M1000MMin: 1052600000 / Avg: 1056400000 / Max: 1061600000Min: 1061000000 / Avg: 1061900000 / Max: 1062700000Min: 1067600000 / Avg: 1073466666.67 / Max: 1078700000Min: 1034000000 / Avg: 1041200000 / Max: 1049900000Min: 1083600000 / Avg: 1085000000 / Max: 1087500000Min: 1092000000 / Avg: 1093266666.67 / Max: 10947000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

libjpeg-turbo tjbench

tjbench is a JPEG decompression/compression benchmark that is part of libjpeg-turbo, a JPEG image codec library optimized for SIMD instructions on modern CPU architectures. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgMegapixels/sec, More Is Betterlibjpeg-turbo tjbench 2.1.0Test: Decompression Throughput-O2-O3 -march=native-O3 -march=native -flto60120180240300SE +/- 1.65, N = 3SE +/- 0.59, N = 3SE +/- 0.54, N = 3SE +/- 1.92, N = 3SE +/- 1.27, N = 3SE +/- 0.45, N = 3267.55272.51266.23277.50268.11270.911. (CC) gcc options: -O3 -rdynamic
Clang 12GCC 11.1OpenBenchmarking.orgMegapixels/sec, More Is Betterlibjpeg-turbo tjbench 2.1.0Test: Decompression Throughput-O2-O3 -march=native-O3 -march=native -flto50100150200250Min: 265.7 / Avg: 267.55 / Max: 270.85Min: 271.91 / Avg: 272.51 / Max: 273.7Min: 265.18 / Avg: 266.23 / Max: 267.01Min: 273.74 / Avg: 277.5 / Max: 280.02Min: 266.22 / Avg: 268.11 / Max: 270.53Min: 270.05 / Avg: 270.91 / Max: 271.551. (CC) gcc options: -O3 -rdynamic

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Medium-O2-O3 -march=native-O3 -march=native -flto1.05742.11483.17224.22965.287SE +/- 0.0156, N = 3SE +/- 0.0243, N = 3SE +/- 0.0302, N = 3SE +/- 0.0337, N = 3SE +/- 0.0292, N = 3SE +/- 0.0045, N = 33.33673.31793.37284.62964.62774.69951. (CXX) g++ options: -O2 -flto -pthread
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Medium-O2-O3 -march=native-O3 -march=native -flto246810Min: 3.31 / Avg: 3.34 / Max: 3.36Min: 3.29 / Avg: 3.32 / Max: 3.37Min: 3.31 / Avg: 3.37 / Max: 3.41Min: 4.59 / Avg: 4.63 / Max: 4.7Min: 4.57 / Avg: 4.63 / Max: 4.67Min: 4.69 / Avg: 4.7 / Max: 4.711. (CXX) g++ options: -O2 -flto -pthread

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Thorough-O2-O3 -march=native-O3 -march=native -flto3691215SE +/- 0.0279, N = 3SE +/- 0.0407, N = 3SE +/- 0.0114, N = 3SE +/- 0.0109, N = 3SE +/- 0.0351, N = 3SE +/- 0.0109, N = 39.39509.35499.44327.68417.57187.52991. (CXX) g++ options: -O2 -flto -pthread
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Thorough-O2-O3 -march=native-O3 -march=native -flto3691215Min: 9.36 / Avg: 9.39 / Max: 9.45Min: 9.28 / Avg: 9.35 / Max: 9.41Min: 9.42 / Avg: 9.44 / Max: 9.46Min: 7.67 / Avg: 7.68 / Max: 7.7Min: 7.5 / Avg: 7.57 / Max: 7.61Min: 7.51 / Avg: 7.53 / Max: 7.551. (CXX) g++ options: -O2 -flto -pthread

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Exhaustive-O2-O3 -march=native-O3 -march=native -flto1326395265SE +/- 0.11, N = 3SE +/- 0.11, N = 3SE +/- 0.07, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.10, N = 352.2051.8251.9057.1456.6456.191. (CXX) g++ options: -O2 -flto -pthread
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterASTC Encoder 2.4Preset: Exhaustive-O2-O3 -march=native-O3 -march=native -flto1122334455Min: 51.99 / Avg: 52.2 / Max: 52.35Min: 51.61 / Avg: 51.82 / Max: 51.96Min: 51.75 / Avg: 51.9 / Max: 51.98Min: 56.97 / Avg: 57.14 / Max: 57.28Min: 56.48 / Avg: 56.64 / Max: 56.73Min: 56.01 / Avg: 56.19 / Max: 56.331. (CXX) g++ options: -O2 -flto -pthread

SQLite Speedtest

This is a benchmark of SQLite's speedtest1 benchmark program with an increased problem size of 1,000. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000-O2-O3 -march=native-O3 -march=native -flto1122334455SE +/- 0.13, N = 3SE +/- 0.50, N = 3SE +/- 0.18, N = 3SE +/- 0.02, N = 3SE +/- 0.16, N = 3SE +/- 0.54, N = 347.7947.6348.5847.4046.9246.481. (CC) gcc options: -ldl -lz -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000-O2-O3 -march=native-O3 -march=native -flto1020304050Min: 47.56 / Avg: 47.79 / Max: 48.01Min: 46.63 / Avg: 47.63 / Max: 48.2Min: 48.28 / Avg: 48.58 / Max: 48.89Min: 47.37 / Avg: 47.4 / Max: 47.44Min: 46.66 / Avg: 46.92 / Max: 47.21Min: 45.41 / Avg: 46.48 / Max: 47.121. (CC) gcc options: -ldl -lz -lpthread

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mobilenet-O2-O3 -march=native-O3 -march=native -flto48121620SE +/- 0.03, N = 3SE +/- 0.11, N = 3SE +/- 0.12, N = 3SE +/- 0.11, N = 15SE +/- 0.13, N = 15SE +/- 0.17, N = 311.5811.4611.0512.9512.7013.741. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mobilenet-O2-O3 -march=native-O3 -march=native -flto48121620Min: 11.53 / Avg: 11.58 / Max: 11.64Min: 11.34 / Avg: 11.46 / Max: 11.68Min: 10.82 / Avg: 11.05 / Max: 11.18Min: 12.28 / Avg: 12.95 / Max: 13.68Min: 12.05 / Avg: 12.7 / Max: 14.02Min: 13.51 / Avg: 13.74 / Max: 14.071. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v2-v2 - Model: mobilenet-v2-O2-O3 -march=native-O3 -march=native -flto1.02152.0433.06454.0865.1075SE +/- 0.11, N = 3SE +/- 0.04, N = 3SE +/- 0.08, N = 3SE +/- 0.04, N = 15SE +/- 0.01, N = 15SE +/- 0.01, N = 33.563.573.424.544.434.461. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v2-v2 - Model: mobilenet-v2-O2-O3 -march=native-O3 -march=native -flto246810Min: 3.42 / Avg: 3.56 / Max: 3.77Min: 3.49 / Avg: 3.57 / Max: 3.64Min: 3.27 / Avg: 3.42 / Max: 3.51Min: 4.45 / Avg: 4.54 / Max: 5.09Min: 4.39 / Avg: 4.43 / Max: 4.51Min: 4.44 / Avg: 4.46 / Max: 4.491. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v3-v3 - Model: mobilenet-v3-O2-O3 -march=native-O3 -march=native -flto0.93381.86762.80143.73524.669SE +/- 0.07, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 15SE +/- 0.02, N = 15SE +/- 0.03, N = 33.363.072.934.153.843.861. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v3-v3 - Model: mobilenet-v3-O2-O3 -march=native-O3 -march=native -flto246810Min: 3.29 / Avg: 3.36 / Max: 3.5Min: 3 / Avg: 3.07 / Max: 3.16Min: 2.92 / Avg: 2.93 / Max: 2.94Min: 4.06 / Avg: 4.15 / Max: 4.42Min: 3.79 / Avg: 3.84 / Max: 4.05Min: 3.81 / Avg: 3.86 / Max: 3.911. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mnasnet-O2-O3 -march=native-O3 -march=native -flto0.89331.78662.67993.57324.4665SE +/- 0.08, N = 3SE +/- 0.05, N = 3SE +/- 0.04, N = 3SE +/- 0.04, N = 15SE +/- 0.04, N = 15SE +/- 0.02, N = 33.263.263.083.973.953.921. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mnasnet-O2-O3 -march=native-O3 -march=native -flto246810Min: 3.18 / Avg: 3.26 / Max: 3.42Min: 3.18 / Avg: 3.26 / Max: 3.36Min: 3.02 / Avg: 3.08 / Max: 3.16Min: 3.88 / Avg: 3.97 / Max: 4.48Min: 3.85 / Avg: 3.95 / Max: 4.35Min: 3.88 / Avg: 3.92 / Max: 3.951. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: efficientnet-b0-O2-O3 -march=native-O3 -march=native -flto1.21732.43463.65194.86926.0865SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 15SE +/- 0.03, N = 15SE +/- 0.02, N = 34.524.514.305.415.375.361. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: efficientnet-b0-O2-O3 -march=native-O3 -march=native -flto246810Min: 4.41 / Avg: 4.52 / Max: 4.63Min: 4.49 / Avg: 4.51 / Max: 4.53Min: 4.25 / Avg: 4.3 / Max: 4.36Min: 5.29 / Avg: 5.41 / Max: 6.01Min: 5.29 / Avg: 5.37 / Max: 5.68Min: 5.32 / Avg: 5.36 / Max: 5.391. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: googlenet-O2-O3 -march=native-O3 -march=native -flto3691215SE +/- 0.13, N = 3SE +/- 0.16, N = 3SE +/- 0.09, N = 3SE +/- 0.22, N = 15SE +/- 0.12, N = 15SE +/- 0.44, N = 311.8811.7711.6213.1512.8513.261. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: googlenet-O2-O3 -march=native-O3 -march=native -flto48121620Min: 11.65 / Avg: 11.88 / Max: 12.1Min: 11.54 / Avg: 11.77 / Max: 12.07Min: 11.45 / Avg: 11.62 / Max: 11.77Min: 12.43 / Avg: 13.15 / Max: 14.57Min: 12.38 / Avg: 12.85 / Max: 13.81Min: 12.7 / Avg: 13.26 / Max: 14.121. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: vgg16-O2-O3 -march=native-O3 -march=native -flto1326395265SE +/- 0.08, N = 3SE +/- 0.53, N = 3SE +/- 0.45, N = 3SE +/- 0.08, N = 15SE +/- 0.08, N = 15SE +/- 0.23, N = 357.2656.9356.1156.8257.6157.841. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: vgg16-O2-O3 -march=native-O3 -march=native -flto1122334455Min: 57.1 / Avg: 57.26 / Max: 57.38Min: 56.12 / Avg: 56.93 / Max: 57.92Min: 55.54 / Avg: 56.11 / Max: 57Min: 56.47 / Avg: 56.82 / Max: 57.5Min: 57.19 / Avg: 57.61 / Max: 58.13Min: 57.45 / Avg: 57.84 / Max: 58.241. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet18-O2-O3 -march=native-O3 -march=native -flto48121620SE +/- 0.23, N = 3SE +/- 0.13, N = 3SE +/- 0.13, N = 3SE +/- 0.11, N = 15SE +/- 0.09, N = 15SE +/- 0.22, N = 314.0513.9013.8614.5014.3314.801. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet18-O2-O3 -march=native-O3 -march=native -flto48121620Min: 13.61 / Avg: 14.05 / Max: 14.36Min: 13.64 / Avg: 13.9 / Max: 14.05Min: 13.62 / Avg: 13.86 / Max: 14.08Min: 13.98 / Avg: 14.5 / Max: 15.18Min: 13.85 / Avg: 14.33 / Max: 14.88Min: 14.36 / Avg: 14.8 / Max: 15.061. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: alexnet-O2-O3 -march=native-O3 -march=native -flto3691215SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.10, N = 3SE +/- 0.06, N = 15SE +/- 0.04, N = 15SE +/- 0.06, N = 311.0911.1611.2511.1610.9911.131. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: alexnet-O2-O3 -march=native-O3 -march=native -flto3691215Min: 11.06 / Avg: 11.09 / Max: 11.15Min: 11.09 / Avg: 11.16 / Max: 11.26Min: 11.08 / Avg: 11.25 / Max: 11.44Min: 10.73 / Avg: 11.16 / Max: 11.53Min: 10.67 / Avg: 10.99 / Max: 11.3Min: 11.03 / Avg: 11.13 / Max: 11.251. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet50-O2-O3 -march=native-O3 -march=native -flto612182430SE +/- 0.11, N = 3SE +/- 0.06, N = 3SE +/- 0.09, N = 3SE +/- 0.18, N = 15SE +/- 0.10, N = 15SE +/- 0.21, N = 322.9923.2522.9026.1825.3325.841. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet50-O2-O3 -march=native-O3 -march=native -flto612182430Min: 22.79 / Avg: 22.99 / Max: 23.18Min: 23.14 / Avg: 23.25 / Max: 23.34Min: 22.79 / Avg: 22.9 / Max: 23.07Min: 24.98 / Avg: 26.18 / Max: 27.12Min: 24.76 / Avg: 25.33 / Max: 25.86Min: 25.42 / Avg: 25.84 / Max: 26.11. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: squeezenet_ssd-O2-O3 -march=native-O3 -march=native -flto48121620SE +/- 0.31, N = 3SE +/- 0.20, N = 3SE +/- 0.20, N = 3SE +/- 0.10, N = 15SE +/- 0.09, N = 15SE +/- 0.06, N = 312.6112.3512.4314.6113.9314.501. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: squeezenet_ssd-O2-O3 -march=native-O3 -march=native -flto48121620Min: 12.27 / Avg: 12.61 / Max: 13.22Min: 12.13 / Avg: 12.35 / Max: 12.75Min: 12.05 / Avg: 12.43 / Max: 12.73Min: 13.63 / Avg: 14.61 / Max: 15.03Min: 13.16 / Avg: 13.93 / Max: 14.57Min: 14.38 / Avg: 14.5 / Max: 14.581. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: regnety_400m-O2-O3 -march=native-O3 -march=native -flto510152025SE +/- 0.20, N = 3SE +/- 0.16, N = 3SE +/- 0.14, N = 3SE +/- 0.06, N = 15SE +/- 0.05, N = 15SE +/- 0.09, N = 312.9312.5912.1417.2417.1818.531. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: regnety_400m-O2-O3 -march=native-O3 -march=native -flto510152025Min: 12.53 / Avg: 12.93 / Max: 13.17Min: 12.26 / Avg: 12.59 / Max: 12.79Min: 11.97 / Avg: 12.14 / Max: 12.42Min: 16.99 / Avg: 17.24 / Max: 17.86Min: 17.05 / Avg: 17.18 / Max: 17.66Min: 18.36 / Avg: 18.53 / Max: 18.651. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

TNN

TNN is an open-source deep learning reasoning framework developed by Tencent. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v2-O2-O3 -march=native-O3 -march=native -flto80160240320400SE +/- 0.41, N = 3SE +/- 0.48, N = 3SE +/- 0.55, N = 3SE +/- 1.57, N = 3SE +/- 0.86, N = 3SE +/- 1.27, N = 3271.03355.93261.84219.13227.77215.341. (CXX) g++ options: -O2 -fopenmp -pthread -fvisibility=hidden -rdynamic -ldl
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: MobileNet v2-O2-O3 -march=native-O3 -march=native -flto60120180240300Min: 270.29 / Avg: 271.03 / Max: 271.72Min: 355 / Avg: 355.93 / Max: 356.62Min: 260.77 / Avg: 261.84 / Max: 262.6Min: 216.8 / Avg: 219.13 / Max: 222.13Min: 226.54 / Avg: 227.77 / Max: 229.43Min: 212.93 / Avg: 215.34 / Max: 217.241. (CXX) g++ options: -O2 -fopenmp -pthread -fvisibility=hidden -rdynamic -ldl

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.1-O2-O3 -march=native-O3 -march=native -flto50100150200250SE +/- 0.23, N = 3SE +/- 0.69, N = 3SE +/- 0.41, N = 3SE +/- 0.08, N = 3SE +/- 0.50, N = 3SE +/- 0.11, N = 3208.56215.78207.93208.78214.28202.571. (CXX) g++ options: -O2 -fopenmp -pthread -fvisibility=hidden -rdynamic -ldl
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterTNN 0.2.3Target: CPU - Model: SqueezeNet v1.1-O2-O3 -march=native-O3 -march=native -flto4080120160200Min: 208.13 / Avg: 208.55 / Max: 208.94Min: 214.47 / Avg: 215.78 / Max: 216.83Min: 207.16 / Avg: 207.93 / Max: 208.55Min: 208.64 / Avg: 208.78 / Max: 208.92Min: 213.29 / Avg: 214.28 / Max: 214.78Min: 202.46 / Avg: 202.57 / Max: 202.791. (CXX) g++ options: -O2 -fopenmp -pthread -fvisibility=hidden -rdynamic -ldl

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: shufflenet-v2-O2-O3 -march=native-O3 -march=native -flto246810SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 14SE +/- 0.15, N = 15SE +/- 0.02, N = 33.813.743.694.334.396.011. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: shufflenet-v2-O2-O3 -march=native-O3 -march=native -flto246810Min: 3.74 / Avg: 3.81 / Max: 3.91Min: 3.67 / Avg: 3.74 / Max: 3.83Min: 3.66 / Avg: 3.69 / Max: 3.73Min: 4.29 / Avg: 4.33 / Max: 4.36Min: 4.2 / Avg: 4.39 / Max: 6.45Min: 5.96 / Avg: 6.01 / Max: 6.031. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: blazeface-O2-O3 -march=native-O3 -march=native -flto0.57381.14761.72142.29522.869SE +/- 0.04, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 15SE +/- 0.01, N = 15SE +/- 0.01, N = 31.541.521.461.851.802.551. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: blazeface-O2-O3 -march=native-O3 -march=native -flto246810Min: 1.49 / Avg: 1.54 / Max: 1.61Min: 1.44 / Avg: 1.52 / Max: 1.58Min: 1.36 / Avg: 1.46 / Max: 1.52Min: 1.81 / Avg: 1.85 / Max: 1.99Min: 1.77 / Avg: 1.8 / Max: 1.87Min: 2.53 / Avg: 2.55 / Max: 2.561. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread

Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: yolov4-tiny-O2-O3 -march=native-O3 -march=native -flto612182430SE +/- 0.18, N = 3SE +/- 0.12, N = 3SE +/- 0.10, N = 3SE +/- 0.11, N = 15SE +/- 0.14, N = 15SE +/- 0.41, N = 321.9721.7321.5821.0621.0023.941. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread
Clang 12GCC 11.1OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: yolov4-tiny-O2-O3 -march=native-O3 -march=native -flto612182430Min: 21.72 / Avg: 21.97 / Max: 22.32Min: 21.5 / Avg: 21.73 / Max: 21.9Min: 21.43 / Avg: 21.58 / Max: 21.78Min: 20.37 / Avg: 21.06 / Max: 21.65Min: 20.34 / Avg: 21 / Max: 22.04Min: 23.12 / Avg: 23.94 / Max: 24.41. (CXX) g++ options: -O2 -rdynamic -lgomp -lpthread