GCC Znver3 First Cut Benchmarks

Tests for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2012046-HA-GCCZNVER333&sgm=1&hgv=znver3&gru&export=txt&rdt&rro.

GCC Znver3 First Cut BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLVulkanCompilerFile-SystemScreen Resolutionznver3x86-64znver1znver2haswellskylakeAMD Ryzen 9 5950X 16-Core @ 3.40GHz (16 Cores / 32 Threads)ASUS ROG CROSSHAIR VIII HERO (WI-FI) (2311 BIOS)AMD Starship/Matisse16GB2000GB Corsair Force MP600 + 2000GBNVIDIA GeForce RTX 3060 Ti 8GB (1665/7000MHz)NVIDIA Device 228bASUS MG28URealtek RTL8125 2.5GbE + Intel I211 + Intel Wi-Fi 6 AX200Ubuntu 20.045.4.0-54-generic (x86_64)GNOME Shell 3.36.4X Server 1.20.8NVIDIA 455.45.014.6.0OpenCL 1.2 CUDA 11.1.1141.2.142GCC 11.0.0 20201203ext43840x2160OpenBenchmarking.orgEnvironment Details- znver3: CXXFLAGS="-O3 -march=znver3" CFLAGS="-O3 -march=znver3"- x86-64: CXXFLAGS="-O3 -march=x86-64" CFLAGS="-O3 -march=x86-64"- znver1: CXXFLAGS="-O3 -march=znver1" CFLAGS="-O3 -march=znver1"- znver2: CXXFLAGS="-O3 -march=znver2" CFLAGS="-O3 -march=znver2"- haswell: CXXFLAGS="-O3 -march=haswell" CFLAGS="-O3 -march=haswell"- skylake: CXXFLAGS="-O3 -march=skylake" CFLAGS="-O3 -march=skylake"Compiler Details- --disable-multilib --enable-checking=releaseDisk Details- znver3: NONE / errors=remount-ro,relatime,rw / Block Size: 4096Processor Details- Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa201009 OpenCL Details- GPU Compute Cores: 4864Python Details- znver3: Python 2.7.18 + Python 3.8.5Security Details- 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

GCC Znver3 First Cut Benchmarksdav1d: Chimera 1080p 10-bitkvazaar: Bosphorus 4K - Ultra Fastkvazaar: Bosphorus 1080p - Ultra Fastvpxenc: Speed 0vpxenc: Speed 5mt-dgemm: Sustained Floating-Point Rategraphics-magick: Sharpengraphics-magick: Enhancedgraphics-magick: Resizingcoremark: CoreMark Size 666 - Iterations Per Secondscimark2: Compositescimark2: Sparse Matrix Multiplyscimark2: Jacobi Successive Over-Relaxationlibraw: Post-Processing Benchmarkcrafty: Elapsed Timehint: FLOATdaphne: OpenMP - NDT Mappingdaphne: OpenMP - Points2Imagewebp: Quality 100c-ray: Total Time - 4K, 16 Rays Per Pixelsmallpt: Global Illumination Renderer; 128 Samplesaobench: 2048 x 2048 - Total Timeencode-mp3: WAV To MP3rnnoise: sqlite-speedtest: Timed Time - Size 1,000znver3x86-64znver1znver2haswellskylake263.9755.82203.0810.7639.138.9723473744472025837788.0842344463.425091.123072.0476.0412060994539663428.10785969.3829625.9578027101.67025.2024.62725.2585.46813.98941.905189.6255.07202.4310.1736.038.0541802253821809783951.8672503662.754977.772440.1065.4811828707521237587.67053932.2929502.0845242881.71731.2125.05929.5215.82214.63342.186251.1156.22205.7110.8238.958.7783163674462084835605.8739753304.214392.612913.8374.5411865359523840766.06954950.3029918.6996710601.69226.0734.79025.0285.54614.09542.250265.1655.85201.8410.7738.978.9221333734462050834342.4401894310.954918.892931.7876.0111910299529343739.09606963.5329824.3403874061.68925.3214.66425.2545.47813.85442.086175.3657.53209.1910.8139.187.8293192995242023776752.2175284332.885062.072934.4674.1311935854519167343.77856969.6029785.6138137081.69025.4554.90625.3285.52015.51242.229262.4055.75203.6910.7738.878.1114863774362023823512.8864064314.574955.382936.0475.0411943425520771613.78862962.7330279.7711343161.68425.0854.87225.3695.56515.57542.783OpenBenchmarking.org

dav1d

Video Input: Chimera 1080p 10-bit

OpenBenchmarking.orgFPS, More Is Betterdav1d 0.7.0Video Input: Chimera 1080p 10-bitskylakehaswellznver2znver1x86-64znver360120180240300SE +/- 0.56, N = 3SE +/- 0.22, N = 3SE +/- 0.73, N = 3SE +/- 0.45, N = 3SE +/- 0.23, N = 3SE +/- 0.85, N = 3262.40175.36265.16251.11189.62263.97-march=skylake - MIN: 173.7 / MAX: 495.13-march=haswell - MIN: 93.55 / MAX: 459.71-march=znver2 - MIN: 175.81 / MAX: 502.41-march=znver1 - MIN: 167.21 / MAX: 471.36-march=x86-64 - MIN: 124.1 / MAX: 372.85-march=znver3 - MIN: 175.34 / MAX: 478.581. (CC) gcc options: -O3 -pthread

Kvazaar

Video Input: Bosphorus 4K - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 4K - Video Preset: Ultra Fastskylakehaswellznver2znver1x86-64znver31326395265SE +/- 0.22, N = 3SE +/- 0.01, N = 3SE +/- 0.10, N = 3SE +/- 0.20, N = 3SE +/- 0.10, N = 3SE +/- 0.09, N = 355.7557.5355.8556.2255.0755.82-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -lpthread -lm -lrt

Kvazaar

Video Input: Bosphorus 1080p - Video Preset: Ultra Fast

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.0Video Input: Bosphorus 1080p - Video Preset: Ultra Fastskylakehaswellznver2znver1x86-64znver350100150200250SE +/- 0.72, N = 3SE +/- 1.14, N = 3SE +/- 1.10, N = 3SE +/- 0.58, N = 3SE +/- 0.29, N = 3SE +/- 0.81, N = 3203.69209.19201.84205.71202.43203.08-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O3 -lpthread -lm -lrt

VP9 libvpx Encoding

Speed: Speed 0

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.8.2Speed: Speed 0skylakehaswellznver2znver1x86-64znver33691215SE +/- 0.02, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 310.7710.8110.7710.8210.1710.76-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=c++11

VP9 libvpx Encoding

Speed: Speed 5

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.8.2Speed: Speed 5skylakehaswellznver2znver1x86-64znver3918273645SE +/- 0.27, N = 3SE +/- 0.02, N = 3SE +/- 0.15, N = 3SE +/- 0.33, N = 3SE +/- 0.09, N = 3SE +/- 0.03, N = 338.8739.1838.9738.9536.0339.13-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=c++11

ACES DGEMM

Sustained Floating-Point Rate

OpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point Rateskylakehaswellznver2znver1x86-64znver33691215SE +/- 0.117999, N = 4SE +/- 0.062576, N = 15SE +/- 0.094255, N = 3SE +/- 0.028561, N = 3SE +/- 0.072954, N = 3SE +/- 0.135072, N = 38.1114867.8293198.9221338.7783168.0541808.972347-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -march=native -fopenmp

GraphicsMagick

Operation: Sharpen

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Sharpenskylakehaswellznver2znver1x86-64znver380160240320400SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.67, N = 3SE +/- 1.00, N = 3377299373367225374-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -fopenmp -O3 -pthread -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

GraphicsMagick

Operation: Enhanced

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Enhancedskylakehaswellznver2znver1x86-64znver3110220330440550SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.33, N = 3SE +/- 0.58, N = 3SE +/- 0.58, N = 3436524446446382447-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -fopenmp -O3 -pthread -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

GraphicsMagick

Operation: Resizing

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.33Operation: Resizingskylakehaswellznver2znver1x86-64znver3400800120016002000SE +/- 4.10, N = 3SE +/- 1.00, N = 3SE +/- 4.91, N = 3SE +/- 0.88, N = 3SE +/- 3.79, N = 3SE +/- 4.33, N = 3202320232050208418092025-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -fopenmp -O3 -pthread -ljbig -lwebp -lwebpmux -ltiff -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -llzma -lbz2 -lxml2 -lz -lm -lpthread

Coremark

CoreMark Size 666 - Iterations Per Second

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per Secondskylakehaswellznver2znver1x86-64znver3200K400K600K800K1000KSE +/- 1894.03, N = 3SE +/- 563.60, N = 3SE +/- 7614.04, N = 3SE +/- 2226.29, N = 3SE +/- 651.07, N = 3SE +/- 5296.50, N = 3823512.89776752.22834342.44835605.87783951.87837788.08-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O2 -O3 -lrt" -lrt

SciMark

Computational Test: Composite

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Compositeskylakehaswellznver2znver1x86-64znver310002000300040005000SE +/- 4.37, N = 3SE +/- 11.48, N = 3SE +/- 1.02, N = 3SE +/- 13.80, N = 3SE +/- 5.93, N = 3SE +/- 0.19, N = 34314.574332.884310.953304.213662.754463.42-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -lm

SciMark

Computational Test: Sparse Matrix Multiply

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Sparse Matrix Multiplyskylakehaswellznver2znver1x86-64znver311002200330044005500SE +/- 10.56, N = 3SE +/- 44.33, N = 3SE +/- 1.75, N = 3SE +/- 1.53, N = 3SE +/- 10.14, N = 3SE +/- 2.08, N = 34955.385062.074918.894392.614977.775091.12-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -lm

SciMark

Computational Test: Jacobi Successive Over-Relaxation

OpenBenchmarking.orgMflops, More Is BetterSciMark 2.0Computational Test: Jacobi Successive Over-Relaxationskylakehaswellznver2znver1x86-64znver37001400210028003500SE +/- 0.45, N = 3SE +/- 2.02, N = 3SE +/- 3.83, N = 3SE +/- 3.35, N = 3SE +/- 0.05, N = 3SE +/- 0.78, N = 32936.042934.462931.782913.832440.103072.04-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -lm

LibRaw

Post-Processing Benchmark

OpenBenchmarking.orgMpix/sec, More Is BetterLibRaw 0.20Post-Processing Benchmarkskylakehaswellznver2znver1x86-64znver320406080100SE +/- 0.39, N = 3SE +/- 0.11, N = 3SE +/- 0.20, N = 3SE +/- 0.16, N = 3SE +/- 0.02, N = 3SE +/- 0.06, N = 375.0474.1376.0174.5465.4876.04-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CXX) g++ options: -O3 -fopenmp -ljpeg -lz -lm

Crafty

Elapsed Time

OpenBenchmarking.orgNodes Per Second, More Is BetterCrafty 25.2Elapsed Timeskylakehaswellznver2znver1x86-64znver33M6M9M12M15MSE +/- 69563.68, N = 3SE +/- 22104.24, N = 3SE +/- 44223.61, N = 3SE +/- 33532.81, N = 3SE +/- 18592.10, N = 3SE +/- 121191.73, N = 31194342511935854119102991186535911828707120609941. (CC) gcc options: -pthread -lstdc++ -fprofile-use -lm

Hierarchical INTegration

Test: FLOAT

OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOATskylakehaswellznver2znver1x86-64znver3120M240M360M480M600MSE +/- 4152069.13, N = 3SE +/- 2024395.89, N = 3SE +/- 5908890.92, N = 3SE +/- 114637.70, N = 3SE +/- 1444160.80, N = 3SE +/- 963540.42, N = 3520771613.79519167343.78529343739.10523840766.07521237587.67539663428.11-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -march=native -lm

Darmstadt Automotive Parallel Heterogeneous Suite

Backend: OpenMP - Kernel: NDT Mapping

OpenBenchmarking.orgTest Cases Per Minute, More Is BetterDarmstadt Automotive Parallel Heterogeneous SuiteBackend: OpenMP - Kernel: NDT Mappingskylakehaswellznver2znver1x86-64znver32004006008001000SE +/- 1.88, N = 3SE +/- 1.90, N = 3SE +/- 2.52, N = 3SE +/- 0.66, N = 3SE +/- 2.69, N = 3SE +/- 3.90, N = 3962.73969.60963.53950.30932.29969.381. (CXX) g++ options: -O3 -std=c++11 -fopenmp

Darmstadt Automotive Parallel Heterogeneous Suite

Backend: OpenMP - Kernel: Points2Image

OpenBenchmarking.orgTest Cases Per Minute, More Is BetterDarmstadt Automotive Parallel Heterogeneous SuiteBackend: OpenMP - Kernel: Points2Imageskylakehaswellznver2znver1x86-64znver36K12K18K24K30KSE +/- 113.76, N = 3SE +/- 289.82, N = 3SE +/- 240.50, N = 3SE +/- 159.95, N = 3SE +/- 108.90, N = 3SE +/- 109.65, N = 330279.7729785.6129824.3429918.7029502.0829625.961. (CXX) g++ options: -O3 -std=c++11 -fopenmp

WebP Image Encode

Encode Settings: Quality 100

OpenBenchmarking.orgEncode Time - Seconds, Fewer Is BetterWebP Image Encode 1.1Encode Settings: Quality 100skylakehaswellznver2znver1x86-64znver30.38630.77261.15891.54521.9315SE +/- 0.001, N = 3SE +/- 0.004, N = 3SE +/- 0.003, N = 3SE +/- 0.004, N = 3SE +/- 0.004, N = 3SE +/- 0.020, N = 31.6841.6901.6891.6921.7171.670-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -fvisibility=hidden -O3 -pthread -lm -ljpeg -lpng16 -ltiff

C-Ray

Total Time - 4K, 16 Rays Per Pixel

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per Pixelskylakehaswellznver2znver1x86-64znver3714212835SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 3SE +/- 0.09, N = 3SE +/- 0.08, N = 3SE +/- 0.07, N = 325.0925.4625.3226.0731.2125.20-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -lm -lpthread -O3

Smallpt

Global Illumination Renderer; 128 Samples

OpenBenchmarking.orgSeconds, Fewer Is BetterSmallpt 1.0Global Illumination Renderer; 128 Samplesskylakehaswellznver2znver1x86-64znver31.13832.27663.41494.55325.6915SE +/- 0.005, N = 3SE +/- 0.014, N = 3SE +/- 0.007, N = 3SE +/- 0.007, N = 3SE +/- 0.010, N = 3SE +/- 0.005, N = 34.8724.9064.6644.7905.0594.627-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CXX) g++ options: -fopenmp -O3

AOBench

Size: 2048 x 2048 - Total Time

OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total Timeskylakehaswellznver2znver1x86-64znver3714212835SE +/- 0.03, N = 3SE +/- 0.10, N = 3SE +/- 0.01, N = 3SE +/- 0.24, N = 3SE +/- 0.68, N = 15SE +/- 0.06, N = 325.3725.3325.2525.0329.5225.26-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -lm -O3

LAME MP3 Encoding

WAV To MP3

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3skylakehaswellznver2znver1x86-64znver31.312.623.935.246.55SE +/- 0.007, N = 3SE +/- 0.020, N = 3SE +/- 0.010, N = 3SE +/- 0.009, N = 3SE +/- 0.015, N = 3SE +/- 0.006, N = 35.5655.5205.4785.5465.8225.468-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr -pipe -lncurses -lm

RNNoise

OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 2020-06-28skylakehaswellznver2znver1x86-64znver348121620SE +/- 0.05, N = 3SE +/- 0.16, N = 3SE +/- 0.19, N = 3SE +/- 0.10, N = 3SE +/- 0.22, N = 3SE +/- 0.13, N = 315.5815.5113.8514.1014.6313.99-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -pedantic -fvisibility=hidden

SQLite Speedtest

Timed Time - Size 1,000

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000skylakehaswellznver2znver1x86-64znver31020304050SE +/- 0.15, N = 3SE +/- 0.23, N = 3SE +/- 0.61, N = 3SE +/- 0.19, N = 3SE +/- 0.18, N = 3SE +/- 0.25, N = 342.7842.2342.0942.2542.1941.91-march=skylake-march=haswell-march=znver2-march=znver1-march=x86-64-march=znver31. (CC) gcc options: -O3 -ldl -lz -lpthread

Geometric Mean Of All Test Results

Result Composite - GCC Znver3 First Cut Benchmarks

OpenBenchmarking.orgGeometric Mean, More Is BetterGeometric Mean Of All Test ResultsResult Composite - GCC Znver3 First Cut Benchmarksskylakehaswellznver2znver1x86-64znver3110220330440550482.49473.43488.84478.18441.20492.26

Phoronix Test Suite v10.8.5