Windows vs. Linux OpenCL - AMD vs. NVIDIA

Windows and Linux OpenCL testing for a future article.

HTML result view exported from: https://openbenchmarking.org/result/1803210-FO-GPGPU256376&rdt.

ProcessorMotherboardMemoryDiskGraphicsNetworkChipsetAudioMonitorOSKernelDisplay DriverOpenCLCompilerFile-SystemScreen ResolutionDesktopOpenGLVulkanWindowsUbuntu RX Vega 64 RX 580 GTX 1080 Ti GTX 1060 GTX 1060 GTX 1080 Ti RX 580 RX Vega 64Intel Core i9-7980XE @ 2.59GHz (18 Cores / 36 Threads)ASUS PRIME X299-A4 x 4096 MB 3200MHz Corsair CMK16GX4M4C3200C16238GB INTEL SSDPEKKW256G8AMD Radeon RX Vega 4095MBIntel Connection I219-VMicrosoft Windows 10 Pro Build 1629910.0 (x86_64)23.20.15033.1003OpenCL 2.1 AMD-APP (2527.10)GCC 7.1.0 + Clang 6.0.0NTFS3840x2160AMD Radeon RX 580 4095MBNVIDIA GeForce GTX 1080 Ti 11264MB (1911/5505MHz)391.24 (23.21.13.9124)OpenCL 1.2 CUDA 9.1.84NVIDIA GeForce GTX 1060 6GB 6144MB (1911/4004MHz)Intel Core i9-7980XE @ 4.40GHz (18 Cores / 36 Threads)ASUS PRIME X299-A (1004 BIOS)Intel Device 202016384MB525GB Crucial_CT525MX3 + 256GB INTEL SSDPEKKW256G8NVIDIA GeForce GTX 1060 6GB 6144MB (1506/4006MHz)Realtek ALC1220Acer B286HKIntel ConnectionUbuntu 16.044.13.0-37-generic (x86_64)Unity 7.4.5NVIDIA 390.424.5.01.0.61GCC 5.4.0 20160609ext4Intel Sky Lake-E DMI3 RegistersNVIDIA GeForce GTX 1080 Ti 11264MB (1480/5508MHz)MSI AMD Radeon RX 580 8192MBmodesetting 1.19.54.5 Mesa 18.1.0-devel- padoka PPA (LLVM 7.0.0)OpenCL 2.1 AMD-APP.internal (2576.0)AMD Radeon RX Vega 8192MBamdgpu 1.4.0OpenBenchmarking.orgCompiler Details- Ubuntu: GTX 1060, Ubuntu: GTX 1080 Ti, Ubuntu: RX 580, Ubuntu: RX Vega 64: --build=x86_64-linux-gnu --disable-browser-plugin --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-gnu-unique-object --enable-gtk-cairo --enable-java-awt=gtk --enable-java-home --enable-languages=c,ada,c++,java,go,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-arch-directory=amd64 --with-default-libstdcxx-abi=new --with-multilib-list=m32,m64,mx32 --with-tune=generic -v Processor Details- Ubuntu: GTX 1060, Ubuntu: GTX 1080 Ti, Ubuntu: RX 580, Ubuntu: RX Vega 64: Scaling Governor: intel_pstate performanceOpenCL Details- Ubuntu: GTX 1060: GPU Compute Cores: 1280- Ubuntu: GTX 1080 Ti: GPU Compute Cores: 3584Security Details- Ubuntu: GTX 1060, Ubuntu: GTX 1080 Ti, Ubuntu: RX 580, Ubuntu: RX Vega 64: KPTI + OSB (observable speculation barrier Intel v6) + Full generic retpoline Protection

darktable: Boat - OpenCLdarktable: Masskrug - OpenCLdarktable: Server Room - OpenCLluxmark: GPU - Hotelluxmark: GPU - Microphoneluxmark: GPU - Luxball HDRdarktable: Boat - OpenCLdarktable: Masskrug - OpenCLdarktable: Server Room - OpenCLWindowsUbuntu RX Vega 64 RX 580 GTX 1080 Ti GTX 1060 GTX 1060 GTX 1080 Ti RX 580 RX Vega 640.750.890.90520617615243580.760.890.8929079850148910.750.890.90572313043205890.780.900.90263668281215026436840121323.820.230.23572613032205702.410.230.231453112421498911.770.210.211898207722520211.840.220.21OpenBenchmarking.org

Darktable

Test: Boat - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.1Test: Boat - Acceleration: OpenCLRX Vega 64RX 580GTX 1080 TiGTX 10600.17550.3510.52650.7020.8775SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 30.750.760.750.78

Darktable

Test: Masskrug - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.1Test: Masskrug - Acceleration: OpenCLRX Vega 64RX 580GTX 1080 TiGTX 10600.20250.4050.60750.811.0125SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.890.890.890.90

Darktable

Test: Server Room - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.1Test: Server Room - Acceleration: OpenCLRX Vega 64RX 580GTX 1080 TiGTX 10600.20250.4050.60750.811.0125SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.00, N = 30.900.890.900.90

LuxMark

OpenCL Device: GPU - Scene: Hotel

WindowsUbuntuOpenBenchmarking.orgScore, More Is BetterLuxMark 3.1OpenCL Device: GPU - Scene: HotelRX Vega 64RX 580GTX 1080 TiGTX 106012002400360048006000SE +/- 36.31, N = 3SE +/- 4.41, N = 3SE +/- 9.54, N = 3SE +/- 12.98, N = 3SE +/- 10.17, N = 3SE +/- 18.50, N = 3SE +/- 5.67, N = 3SE +/- 6.96, N = 352062907572326362643572614531898

LuxMark

OpenCL Device: GPU - Scene: Microphone

WindowsUbuntuOpenBenchmarking.orgScore, More Is BetterLuxMark 3.1OpenCL Device: GPU - Scene: MicrophoneRX Vega 64RX 580GTX 1080 TiGTX 10604K8K12K16K20KSE +/- 106.10, N = 3SE +/- 17.61, N = 3SE +/- 29.74, N = 3SE +/- 23.88, N = 3SE +/- 19.38, N = 3SE +/- 20.61, N = 3SE +/- 55.60, N = 3SE +/- 37.29, N = 31761598501304368286840130321124220772

LuxMark

OpenCL Device: GPU - Scene: Luxball HDR

WindowsUbuntuOpenBenchmarking.orgScore, More Is BetterLuxMark 3.1OpenCL Device: GPU - Scene: Luxball HDRRX Vega 64RX 580GTX 1080 TiGTX 10605K10K15K20K25KSE +/- 111.17, N = 3SE +/- 51.12, N = 3SE +/- 38.28, N = 3SE +/- 12.45, N = 3SE +/- 60.70, N = 3SE +/- 69.31, N = 3SE +/- 61.76, N = 3SE +/- 369.64, N = 32435814891205891215012132205701498925202

Darktable

Test: Boat - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.0.3Test: Boat - Acceleration: OpenCLGTX 1060GTX 1080 TiRX 580RX Vega 643691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.36, N = 6SE +/- 0.27, N = 63.822.4111.7711.84

Darktable

Test: Masskrug - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.0.3Test: Masskrug - Acceleration: OpenCLGTX 1060GTX 1080 TiRX 580RX Vega 640.05180.10360.15540.20720.259SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.230.230.210.22

Darktable

Test: Server Room - Acceleration: OpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.0.3Test: Server Room - Acceleration: OpenCLGTX 1060GTX 1080 TiRX 580RX Vega 640.05180.10360.15540.20720.259SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.230.230.210.21


Phoronix Test Suite v10.8.4