Intel OpenCL Linux Ubuntu 19.04 Comparison

Intel Core i7 OpenCL NEO benchmarking for a future article by Michael Larabel.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 1905084-HV-INTELOPEN36
Jump To Table - Results

View

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

Limit displaying results to tests within:

CPU Massive 4 Tests
Creator Workloads 2 Tests
HPC - High Performance Computing 2 Tests
Machine Learning 2 Tests
Multi-Core 2 Tests
NVIDIA GPU Compute 6 Tests
OpenCL 7 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
Beignet OpenCL 2.0
May 07 2019
  2 Hours, 27 Minutes
NEO OpenCL 2.1
May 08 2019
  1 Hour, 47 Minutes
POCL OpenCL 1.2
May 08 2019
  2 Hours
CPU Only
May 08 2019
  2 Hours, 12 Minutes
Invert Hiding All Results Option
  2 Hours, 7 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):


Intel OpenCL Linux Ubuntu 19.04 ComparisonOpenBenchmarking.orgPhoronix Test SuiteIntel Core i7-6770HQ @ 3.50GHz (4 Cores / 8 Threads)Intel NUC6i7KYB (KYSKLi70.86A.0037.2016.0603.1032 BIOS)Intel Xeon E3-1200 v5/E3-150032768MBSamsung SSD 950 PRO 512GBIntel Iris Pro 580 3GB (950MHz)Realtek ALC233DELL P2415QIntel I219-LM + Intel 8260Ubuntu 19.045.0.0-13-generic (x86_64)GNOME Shell 3.32.0X Server 1.20.4modesetting 1.20.44.5 Mesa 19.1.0-devel (git-bdd273d 2019-05-06 disco-oibaf-ppa)OpenCL 2.0 beignet 1.3OpenCL 2.1OpenCL 1.2 pocl +Asserts LLVM 6.0.1 SLEEF DISTRO POCL_DEBUGGCC 8.3.0ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLsCompilerFile-SystemScreen ResolutionIntel OpenCL Linux Ubuntu 19.04 Comparison BenchmarksSystem Logs- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v - Scaling Governor: intel_pstate powersave- KPTI + __user pointer sanitization + Full generic retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + SSB disabled via prctl and seccomp + PTE Inversion; VMX: conditional cache flushes SMT vulnerable

Beignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU OnlyResult OverviewPhoronix Test Suite100%163%226%290%LeelaChessZeroBlenderDarktable

Intel OpenCL Linux Ubuntu 19.04 Comparisonshoc: OpenCL - Triadshoc: OpenCL - FFT SPshoc: OpenCL - MD5 Hashshoc: OpenCL - Max SP Flopsshoc: OpenCL - Bus Speed Downloadshoc: OpenCL - Bus Speed Readbackshoc: OpenCL - Texture Read Bandwidthviennacl: OpenCL LU Factorizationcl-mem: Copycl-mem: Readcl-mem: Writelczero: OpenCLdarktable: Boat - OpenCLdarktable: Masskrug - OpenCLdarktable: Server Rack - OpenCLdarktable: Server Room - OpenCLblender: BMW27 - OpenCLblender: Barbershop - OpenCLxsbench-cl: juliagpu: GPUclpeak: Kernel Latencyclpeak: Single-Precision Floatclpeak: Global Memory Bandwidthclpeak: Transfer Bandwidth enqueueReadBufferclpeak: Transfer Bandwidth enqueueWriteBuffercomd-cl: Average Atom Update Ratelczero: BLASdarktable: Boat - CPU-onlydarktable: Masskrug - CPU-onlydarktable: Server Rack - CPU-onlydarktable: Server Room - CPU-onlyblender: BMW27 - CPU-Onlyblender: Barbershop - CPU-Onlyxsbench: Beignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only8.5428.630.8984510.7318.6214112.9246.6543.0948.9211616.5912.020.288.416952779236249494921872631.25105425.649.3728.943.7317.1841.240.94317627.6327.9312820.2515515.6410.670.525.3067927456687006533.80107725.789.5929.473.6112.212.360.0619015.9415.625.149.8144.6016.5711.980.288.39665273812.08157.3616.5412.010.288.36545.092628.02795093157.3616.5412.010.288.36545.092628.02795093OpenBenchmarking.org

SHOC Scalable HeterOgeneous Computing

The CUDA and OpenCL version of Vetter's Scalable HeterOgeneous Computing benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: TriadBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.248121620SE +/- 0.27, N = 12SE +/- 0.15, N = 3SE +/- 0.06, N = 38.5417.1812.211. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: TriadBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.248121620Min: 7.6 / Avg: 8.54 / Max: 11.04Min: 16.92 / Avg: 17.18 / Max: 17.45Min: 12.09 / Avg: 12.21 / Max: 12.31. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: FFT SPBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2918273645SE +/- 0.01, N = 3SE +/- 0.04, N = 3SE +/- 0.00, N = 328.6341.242.361. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: FFT SPBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2918273645Min: 28.62 / Avg: 28.63 / Max: 28.64Min: 41.2 / Avg: 41.24 / Max: 41.33Min: 2.36 / Avg: 2.36 / Max: 2.361. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: MD5 HashBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.20.21150.4230.63450.8461.0575SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.890.940.061. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: MD5 HashBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2246810Min: 0.89 / Avg: 0.89 / Max: 0.89Min: 0.94 / Avg: 0.94 / Max: 0.94Min: 0.06 / Avg: 0.06 / Max: 0.061. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Max SP FlopsBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.27001400210028003500SE +/- 0.12, N = 3SE +/- 2.05, N = 3SE +/- 2.12, N = 784531761901. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Max SP FlopsBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.26001200180024003000Min: 845.02 / Avg: 845.23 / Max: 845.43Min: 3172.12 / Avg: 3176.21 / Max: 3178.61Min: 177.09 / Avg: 189.55 / Max: 192.341. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed DownloadBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2714212835SE +/- 0.14, N = 4SE +/- 0.39, N = 4SE +/- 0.26, N = 310.7327.6315.941. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed DownloadBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2612182430Min: 10.37 / Avg: 10.73 / Max: 11.04Min: 26.78 / Avg: 27.63 / Max: 28.6Min: 15.58 / Avg: 15.94 / Max: 16.451. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed ReadbackBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2714212835SE +/- 0.15, N = 3SE +/- 0.32, N = 6SE +/- 0.09, N = 318.6227.9315.621. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed ReadbackBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2612182430Min: 18.46 / Avg: 18.62 / Max: 18.91Min: 26.82 / Avg: 27.93 / Max: 28.65Min: 15.5 / Avg: 15.62 / Max: 15.811. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Texture Read BandwidthBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2306090120150SE +/- 0.20, N = 3SE +/- 0.37, N = 3SE +/- 0.00, N = 3141.00128.005.141. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Texture Read BandwidthBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2306090120150Min: 140.77 / Avg: 140.98 / Max: 141.38Min: 127.32 / Avg: 127.78 / Max: 128.52Min: 5.13 / Avg: 5.14 / Max: 5.141. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

ViennaCL

ViennaCL is an open-source linear algebra library written in C++ and with support for OpenCL and OpenMP. This test profile uses ViennaCL OpenCL support and runs the included computational benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS, More Is BetterViennaCL 1.4.2OpenCL LU FactorizationBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2510152025SE +/- 0.61, N = 15SE +/- 0.02, N = 3SE +/- 0.05, N = 312.9220.259.811. (CXX) g++ options: -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterViennaCL 1.4.2OpenCL LU FactorizationBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2510152025Min: 11.17 / Avg: 12.92 / Max: 19.54Min: 20.21 / Avg: 20.25 / Max: 20.29Min: 9.73 / Avg: 9.81 / Max: 9.891. (CXX) g++ options: -rdynamic -lOpenCL

cl-mem

A basic OpenCL memory benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: CopyBeignet OpenCL 2.01122334455SE +/- 2.67, N = 1246.651. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadBeignet OpenCL 2.01020304050SE +/- 2.62, N = 1243.091. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteBeignet OpenCL 2.01122334455SE +/- 1.77, N = 1548.921. (CC) gcc options: -O2 -flto -lOpenCL

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.20.1Backend: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only306090120150SE +/- 6.45, N = 12SE +/- 1.39, N = 12SE +/- 0.05, N = 3SE +/- 0.83, N = 3116.00155.0044.60157.361. (CXX) g++ options: -lpthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.20.1Backend: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only306090120150Min: 71.55 / Avg: 115.79 / Max: 140.67Min: 148.44 / Avg: 154.77 / Max: 167.6Min: 44.51 / Avg: 44.6 / Max: 44.66Min: 155.98 / Avg: 157.36 / Max: 158.841. (CXX) g++ options: -lpthread

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Boat - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only48121620SE +/- 0.04, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3SE +/- 0.03, N = 316.5915.6416.5716.54
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Boat - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only48121620Min: 16.51 / Avg: 16.59 / Max: 16.63Min: 15.52 / Avg: 15.64 / Max: 15.73Min: 16.49 / Avg: 16.57 / Max: 16.63Min: 16.49 / Avg: 16.54 / Max: 16.59

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Masskrug - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only3691215SE +/- 0.01, N = 3SE +/- 0.06, N = 3SE +/- 0.02, N = 3SE +/- 0.02, N = 312.0210.6711.9812.01
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Masskrug - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only48121620Min: 11.99 / Avg: 12.02 / Max: 12.04Min: 10.56 / Avg: 10.67 / Max: 10.76Min: 11.95 / Avg: 11.98 / Max: 12.02Min: 11.98 / Avg: 12.01 / Max: 12.04

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Rack - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only0.1170.2340.3510.4680.585SE +/- 0.00, N = 4SE +/- 0.01, N = 5SE +/- 0.00, N = 3SE +/- 0.00, N = 30.280.520.280.28
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Rack - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only246810Min: 0.28 / Avg: 0.28 / Max: 0.29Min: 0.5 / Avg: 0.52 / Max: 0.54Min: 0.28 / Avg: 0.28 / Max: 0.28Min: 0.28 / Avg: 0.28 / Max: 0.29

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Room - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only246810SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 38.415.308.398.36
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Room - Acceleration: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only3691215Min: 8.41 / Avg: 8.41 / Max: 8.42Min: 5.25 / Avg: 5.3 / Max: 5.34Min: 8.36 / Avg: 8.39 / Max: 8.43Min: 8.35 / Avg: 8.36 / Max: 8.37

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: BMW27 - Compute: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only150300450600750695.00679.00665.00545.09

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Barbershop - Compute: OpenCLBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2CPU Only60012001800240030002779.002745.002738.002628.02

Xsbench OpenCL

Xsbench benchmark in OpenCL via GPUOpen. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgLookups/s, More Is BetterXsbench OpenCL 2017-07-06Beignet OpenCL 2.0CPU Only5M10M15M20M25MSE +/- 12258.51, N = 3SE +/- 2180.82, N = 323624949795093-lOpenCL1. (CC) gcc options: -std=gnu99 -fopenmp -O3 -lm
OpenBenchmarking.orgLookups/s, More Is BetterXsbench OpenCL 2017-07-06Beignet OpenCL 2.0CPU Only4M8M12M16M20MMin: 23602799 / Avg: 23624949 / Max: 23645126Min: 792564 / Avg: 795093 / Max: 7994351. (CC) gcc options: -std=gnu99 -fopenmp -O3 -lm

JuliaGPU

JuliaGPU is an OpenCL benchmark with this version containing various PTS-specific enhancements. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSamples/sec, More Is BetterJuliaGPU 1.2pts1OpenCL Device: GPUBeignet OpenCL 2.0NEO OpenCL 2.114M28M42M56M70MSE +/- 105679.48, N = 3SE +/- 208041.13, N = 349218726668700651. (CC) gcc options: -O3 -march=native -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL -lm
OpenBenchmarking.orgSamples/sec, More Is BetterJuliaGPU 1.2pts1OpenCL Device: GPUBeignet OpenCL 2.0NEO OpenCL 2.112M24M36M48M60MMin: 49011716.1 / Avg: 49218726.37 / Max: 49359170.9Min: 66586377.4 / Avg: 66870065.23 / Max: 67275507.81. (CC) gcc options: -O3 -march=native -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL -lm

clpeak

Clpeak is designed to test the peak capabilities of OpenCL devices. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus, Fewer Is BetterclpeakOpenCL Test: Kernel LatencyBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2816243240SE +/- 0.28, N = 15SE +/- 0.48, N = 3SE +/- 0.02, N = 331.2533.8012.081. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgus, Fewer Is BetterclpeakOpenCL Test: Kernel LatencyBeignet OpenCL 2.0NEO OpenCL 2.1POCL OpenCL 1.2714212835Min: 29.05 / Avg: 31.25 / Max: 32.53Min: 32.86 / Avg: 33.8 / Max: 34.44Min: 12.04 / Avg: 12.08 / Max: 12.121. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatBeignet OpenCL 2.0NEO OpenCL 2.12004006008001000SE +/- 0.23, N = 3SE +/- 3.42, N = 3105410771. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatBeignet OpenCL 2.0NEO OpenCL 2.12004006008001000Min: 1053.97 / Avg: 1054.27 / Max: 1054.72Min: 1073.33 / Avg: 1077.16 / Max: 1083.981. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthBeignet OpenCL 2.0NEO OpenCL 2.1612182430SE +/- 0.16, N = 3SE +/- 0.12, N = 325.6425.781. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthBeignet OpenCL 2.0NEO OpenCL 2.1612182430Min: 25.46 / Avg: 25.64 / Max: 25.96Min: 25.62 / Avg: 25.78 / Max: 26.011. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueReadBufferBeignet OpenCL 2.0NEO OpenCL 2.13691215SE +/- 0.06, N = 3SE +/- 0.06, N = 39.379.591. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueReadBufferBeignet OpenCL 2.0NEO OpenCL 2.13691215Min: 9.31 / Avg: 9.37 / Max: 9.5Min: 9.49 / Avg: 9.59 / Max: 9.71. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueWriteBufferBeignet OpenCL 2.0NEO OpenCL 2.1714212835SE +/- 0.29, N = 3SE +/- 0.14, N = 328.9429.471. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueWriteBufferBeignet OpenCL 2.0NEO OpenCL 2.1714212835Min: 28.62 / Avg: 28.94 / Max: 29.52Min: 29.32 / Avg: 29.47 / Max: 29.761. (CXX) g++ options: -O3 -rdynamic -lOpenCL

CoMD OpenCL

CoMD benchmark in OpenCL via GPUOpen. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus/atom/task, More Is BetterCoMD OpenCL 2017-07-06Average Atom Update RateBeignet OpenCL 2.0NEO OpenCL 2.10.83931.67862.51793.35724.1965SE +/- 0.05, N = 4SE +/- 0.02, N = 33.733.611. (CC) gcc options: -std=c99 -O5 -lm -lOpenCL
OpenBenchmarking.orgus/atom/task, More Is BetterCoMD OpenCL 2017-07-06Average Atom Update RateBeignet OpenCL 2.0NEO OpenCL 2.1246810Min: 3.65 / Avg: 3.73 / Max: 3.88Min: 3.58 / Avg: 3.61 / Max: 3.661. (CC) gcc options: -std=c99 -O5 -lm -lOpenCL

LeelaChessZero

LeelaChessZero (lc0 / lczero) is a chess engine automated vian neural networks. This test profile can be used for OpenCL, CUDA + cuDNN, and BLAS (CPU-based) benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.20.1Backend: BLASCPU Only306090120150SE +/- 0.83, N = 3157.361. (CXX) g++ options: -lpthread

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Boat - Acceleration: CPU-onlyCPU Only48121620SE +/- 0.03, N = 316.54

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Masskrug - Acceleration: CPU-onlyCPU Only3691215SE +/- 0.02, N = 312.01

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Rack - Acceleration: CPU-onlyCPU Only0.0630.1260.1890.2520.315SE +/- 0.00, N = 30.28

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Room - Acceleration: CPU-onlyCPU Only246810SE +/- 0.01, N = 38.36

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: BMW27 - Compute: CPU-OnlyCPU Only120240360480600545.09

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Barbershop - Compute: CPU-OnlyCPU Only60012001800240030002628.02

Xsbench

XSBench is a mini-app representing a key computational kernel of the Monte Carlo neutronics application OpenMC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgLookups/s, More Is BetterXsbench 2017-07-06CPU Only200K400K600K800K1000KSE +/- 2180.82, N = 37950931. (CC) gcc options: -std=gnu99 -fopenmp -O3 -lm