GTX 950 CompuLab Airtop

Intel Core i7-5775C testing with a CompuLab v1.0 (ARTP-3.1.0.637.0.0 X64 BIOS) and eVGA NVIDIA GeForce GTX 950 2GB on Ubuntu 20.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2011049-FI-GTX950COM84
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:

BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
CPU Massive 2 Tests
Creator Workloads 2 Tests
HPC - High Performance Computing 3 Tests
Machine Learning 2 Tests
Multi-Core 3 Tests
NVIDIA GPU Compute 18 Tests
OpenCL 4 Tests
Vulkan Compute 4 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
Run 1
November 01 2020
  20 Hours, 5 Minutes
2
November 03 2020
  14 Hours, 33 Minutes
3
November 03 2020
  14 Hours, 37 Minutes
Invert Hiding All Results Option
  16 Hours, 25 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):


GTX 950 CompuLab AirtopProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLVulkanCompilerFile-SystemScreen ResolutionRun 123Intel Core i7-5775C @ 3.70GHz (4 Cores / 8 Threads)CompuLab v1.0 (ARTP-3.1.0.637.0.0 X64 BIOS)Intel Broadwell-U DMI16GB256GB ADATA SP310eVGA NVIDIA GeForce GTX 950 2GB (1151/3304MHz)Realtek ALC886G237HLIntel I218-LM + Intel I210 + Intel 7260Ubuntu 20.105.8.0-26-generic (x86_64)GNOME Shell 3.38.1X Server 1.20.9NVIDIA 455.284.6.0OpenCL 1.2 CUDA 11.1.961.2.142GCC 10.2.0ext41920x1080OpenBenchmarking.orgCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --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++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --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=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: intel_cpufreq ondemand - CPU Microcode: 0x22 - Thermald 2.3 OpenCL Details- GPU Compute Cores: 768Python Details- Run 1: Python 3.8.6Security Details- itlb_multihit: KVM: Mitigation of VMX disabled + l1tf: Mitigation of PTE Inversion; VMX: conditional cache flushes SMT vulnerable + mds: Mitigation of Clear buffers; SMT vulnerable + meltdown: Mitigation of PTI + 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 generic retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Mitigation of Microcode + tsx_async_abort: Mitigation of Clear buffers; SMT vulnerable

Run 123Result OverviewPhoronix Test Suite 10.6.0100%104%108%111%115%BlenderNAMD CUDAFinanceBenchNCNNOctaneBenchLeelaChessZeroLuxCoreRender OpenCLVkFFTWaifu2x-NCNN VulkanRealSR-NCNNRedShift DemoMandelGPUHashcatArrayFireNeatBenchclpeakFAHBenchViennaCLcl-mem

GTX 950 CompuLab Airtoprealsr-ncnn: 4x - Norealsr-ncnn: 4x - Yeswaifu2x-ncnn: 2x - 3 - Nowaifu2x-ncnn: 2x - 3 - Yesvkfft: hashcat: MD5hashcat: SHA1hashcat: SHA-512financebench: Black-Scholes OpenCLviennacl: OpenCL LU Factorizationcl-mem: Copycl-mem: Readcl-mem: Writenamd-cuda: ATPase Simulation - 327,506 Atomsoctanebench: Total Scoreredshift: luxcorerender-cl: DLSCluxcorerender-cl: Foodluxcorerender-cl: LuxCore Benchmarkluxcorerender-cl: Rainbow Colors and Prismfahbench: lczero: OpenCLarrayfire: Conjugate Gradient OpenCLncnn: Vulkan GPU - squeezenetncnn: Vulkan GPU - mobilenetncnn: Vulkan GPU-v2-v2 - mobilenet-v2ncnn: Vulkan GPU-v3-v3 - mobilenet-v3ncnn: Vulkan GPU - shufflenet-v2ncnn: Vulkan GPU - mnasnetncnn: Vulkan GPU - efficientnet-b0ncnn: Vulkan GPU - blazefacencnn: Vulkan GPU - googlenetncnn: Vulkan GPU - vgg16ncnn: Vulkan GPU - resnet18ncnn: Vulkan GPU - alexnetncnn: Vulkan GPU - resnet50ncnn: Vulkan GPU - yolov4-tinyblender: BMW27 - CUDAblender: Classroom - CUDAblender: Fishy Cat - CUDAblender: Barbershop - CUDAblender: BMW27 - NVIDIA OptiXblender: Classroom - NVIDIA OptiXblender: Fishy Cat - NVIDIA OptiXblender: Pabellon Barcelona - CUDAblender: Pabellon Barcelona - NVIDIA OptiXmandelgpu: GPUclpeak: Integer Compute INTclpeak: Single-Precision Floatclpeak: Double-Precision Doubleclpeak: Global Memory Bandwidthneatbench: GPURun 12347.241404.6533.28020.25166334834000000168763333318190000029.13193335.661979.290.8880.6524543.92840921310.900.320.713.1049.397434448.29612.4410.843.213.772.623.355.781.018.5734.006.248.8114.3420.05308.30894.18646.843167.76349.291021.55715.411896.871960.7154591952.5542.121870.1866.6690.689.5247.240405.5163.28420.26666184835500000168760000017136666726.67666735.597279.290.887.80.6962643.60368121310.930.330.723.1649.760334138.29712.5810.923.223.782.643.365.841.018.5635.216.378.8615.1120.10381.11912.44827.473297.83403.461099.24949.112114.702301.3853723378.2531.801859.8765.6289.329.3948.298413.8123.38020.64064664706166667165943333318403333327.37266735.360079.290.887.80.7138142.38202821760.910.310.713.0949.059933378.42012.9111.193.333.982.803.476.061.108.9036.206.609.0115.7520.04381.65912.76827.393297.57403.511100.13950.932109.622301.0454824518.7532.001842.1265.8389.599.38OpenBenchmarking.org

RealSR-NCNN

RealSR-NCNN is an NCNN neural network implementation of the RealSR project and accelerated using the Vulkan API. RealSR is the Real-World Super Resolution via Kernel Estimation and Noise Injection. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image by a scale of 4x with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoRun 1231122334455SE +/- 0.04, N = 3SE +/- 0.02, N = 3SE +/- 0.07, N = 347.2447.2448.30
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: NoRun 1231020304050Min: 47.19 / Avg: 47.24 / Max: 47.33Min: 47.22 / Avg: 47.24 / Max: 47.27Min: 48.17 / Avg: 48.3 / Max: 48.41

OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesRun 12390180270360450SE +/- 6.37, N = 9SE +/- 6.32, N = 9SE +/- 6.35, N = 9404.65405.52413.81
OpenBenchmarking.orgSeconds, Fewer Is BetterRealSR-NCNN 20200818Scale: 4x - TAA: YesRun 12370140210280350Min: 362.09 / Avg: 404.65 / Max: 417.59Min: 362.96 / Avg: 405.52 / Max: 417.99Min: 370.88 / Avg: 413.81 / Max: 426.98

Waifu2x-NCNN Vulkan

Waifu2x-NCNN is an NCNN neural network implementation of the Waifu2x converter project and accelerated using the Vulkan API. NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. This test profile times how long it takes to increase the resolution of a sample image with Vulkan. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: NoRun 1230.76051.5212.28153.0423.8025SE +/- 0.012, N = 3SE +/- 0.006, N = 3SE +/- 0.013, N = 33.2803.2843.380
OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: NoRun 123246810Min: 3.27 / Avg: 3.28 / Max: 3.3Min: 3.27 / Avg: 3.28 / Max: 3.29Min: 3.37 / Avg: 3.38 / Max: 3.41

OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesRun 123510152025SE +/- 0.10, N = 3SE +/- 0.09, N = 3SE +/- 0.09, N = 320.2520.2720.64
OpenBenchmarking.orgSeconds, Fewer Is BetterWaifu2x-NCNN Vulkan 20200818Scale: 2x - Denoise: 3 - TAA: YesRun 123510152025Min: 20.06 / Avg: 20.25 / Max: 20.39Min: 20.09 / Avg: 20.27 / Max: 20.38Min: 20.45 / Avg: 20.64 / Max: 20.76

VkFFT

VkFFT is a Fast Fourier Transform (FFT) Library that is GPU accelerated by means of the Vulkan API. The VkFFT benchmark runs FFT performance differences of many different sizes before returning an overall benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 2020-09-29Run 12314002800420056007000SE +/- 32.85, N = 3SE +/- 18.01, N = 3SE +/- 16.86, N = 3663366186466
OpenBenchmarking.orgBenchmark Score, More Is BetterVkFFT 2020-09-29Run 12312002400360048006000Min: 6594 / Avg: 6632.67 / Max: 6698Min: 6599 / Avg: 6618 / Max: 6654Min: 6448 / Avg: 6466.33 / Max: 6500

Hashcat

Hashcat is an open-source, advanced password recovery tool supporting GPU acceleration with OpenCL, NVIDIA CUDA, and Radeon ROCm. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: MD5Run 1231000M2000M3000M4000M5000MSE +/- 26534003.34, N = 3SE +/- 8524670.08, N = 3SE +/- 29849195.04, N = 3483400000048355000004706166667
OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: MD5Run 123800M1600M2400M3200M4000MMin: 4793600000 / Avg: 4834000000 / Max: 4884000000Min: 4821600000 / Avg: 4835500000 / Max: 4851000000Min: 4647300000 / Avg: 4706166666.67 / Max: 4744200000

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA1Run 123400M800M1200M1600M2000MSE +/- 6483140.53, N = 3SE +/- 5008991.91, N = 3SE +/- 3933757.04, N = 3168763333316876000001659433333
OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA1Run 123300M600M900M1200M1500MMin: 1674700000 / Avg: 1687633333.33 / Max: 1694900000Min: 1679900000 / Avg: 1687600000 / Max: 1697000000Min: 1651800000 / Avg: 1659433333.33 / Max: 1664900000

OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA-512Run 12340M80M120M160M200MSE +/- 1660177.84, N = 15SE +/- 536449.23, N = 3SE +/- 635959.47, N = 3181900000171366667184033333
OpenBenchmarking.orgH/s, More Is BetterHashcat 6.1.1Benchmark: SHA-512Run 12330M60M90M120M150MMin: 172100000 / Avg: 181900000 / Max: 188500000Min: 170600000 / Avg: 171366666.67 / Max: 172400000Min: 182900000 / Avg: 184033333.33 / Max: 185100000

FinanceBench

FinanceBench is a collection of financial program benchmarks with support for benchmarking on the GPU. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterFinanceBench 2016-06-06Benchmark: Black-Scholes OpenCLRun 123714212835SE +/- 0.38, N = 15SE +/- 0.08, N = 3SE +/- 0.00, N = 329.1326.6827.371. (CXX) g++ options: -O3 -lOpenCL
OpenBenchmarking.orgms, Fewer Is BetterFinanceBench 2016-06-06Benchmark: Black-Scholes OpenCLRun 123612182430Min: 25.5 / Avg: 29.13 / Max: 29.71Min: 26.59 / Avg: 26.68 / Max: 26.85Min: 27.36 / Avg: 27.37 / Max: 27.381. (CXX) g++ options: -O3 -lOpenCL

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 FactorizationRun 123816243240SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 335.6635.6035.361. (CXX) g++ options: -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterViennaCL 1.4.2OpenCL LU FactorizationRun 123816243240Min: 35.65 / Avg: 35.66 / Max: 35.68Min: 35.58 / Avg: 35.6 / Max: 35.61Min: 35.33 / Avg: 35.36 / Max: 35.41. (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: CopyRun 12320406080100SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 379.279.279.21. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: CopyRun 1231530456075Min: 79.2 / Avg: 79.2 / Max: 79.2Min: 79.2 / Avg: 79.2 / Max: 79.2Min: 79.2 / Avg: 79.2 / Max: 79.21. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadRun 12320406080100SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 390.890.890.81. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadRun 12320406080100Min: 90.8 / Avg: 90.8 / Max: 90.8Min: 90.8 / Avg: 90.8 / Max: 90.8Min: 90.8 / Avg: 90.8 / Max: 90.81. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteRun 12320406080100SE +/- 0.03, N = 3SE +/- 0.07, N = 388.087.887.81. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteRun 12320406080100Min: 87.8 / Avg: 87.83 / Max: 87.9Min: 87.7 / Avg: 87.83 / Max: 87.91. (CC) gcc options: -O2 -flto -lOpenCL

NAMD CUDA

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. This version of the NAMD test profile uses CUDA GPU acceleration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD CUDA 2.14ATPase Simulation - 327,506 AtomsRun 1230.16060.32120.48180.64240.803SE +/- 0.00639, N = 3SE +/- 0.00141, N = 3SE +/- 0.00135, N = 30.652450.696260.71381
OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD CUDA 2.14ATPase Simulation - 327,506 AtomsRun 123246810Min: 0.64 / Avg: 0.65 / Max: 0.66Min: 0.69 / Avg: 0.7 / Max: 0.7Min: 0.71 / Avg: 0.71 / Max: 0.72

OctaneBench

OctaneBench is a test of the OctaneRender on the GPU and requires the use of NVIDIA CUDA. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterOctaneBench 2020.1Total ScoreRun 123102030405043.9343.6042.38

RedShift Demo

This is a test of MAXON's RedShift demo build that currently requires NVIDIA GPU acceleration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRedShift Demo 3.0Run 1235001000150020002500SE +/- 2.03, N = 3SE +/- 2.19, N = 3213121312176
OpenBenchmarking.orgSeconds, Fewer Is BetterRedShift Demo 3.0Run 123400800120016002000Min: 2128 / Avg: 2131.33 / Max: 2135Min: 2127 / Avg: 2131.33 / Max: 2134

LuxCoreRender OpenCL

LuxCoreRender is an open-source physically based renderer. This test profile is focused on running LuxCoreRender on OpenCL accelerators/GPUs. The alternative luxcorerender test profile is for CPU execution due to a difference in tests, etc. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: DLSCRun 1230.20930.41860.62790.83721.0465SE +/- 0.02, N = 12SE +/- 0.00, N = 3SE +/- 0.00, N = 30.900.930.91MIN: 0.19 / MAX: 0.94MIN: 0.85 / MAX: 0.95MIN: 0.83 / MAX: 0.93
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: DLSCRun 123246810Min: 0.64 / Avg: 0.9 / Max: 0.93Min: 0.93 / Avg: 0.93 / Max: 0.93Min: 0.91 / Avg: 0.91 / Max: 0.91

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: FoodRun 1230.07430.14860.22290.29720.3715SE +/- 0.01, N = 12SE +/- 0.00, N = 3SE +/- 0.00, N = 30.320.330.31MIN: 0.06 / MAX: 0.42MIN: 0.11 / MAX: 0.41MIN: 0.1 / MAX: 0.4
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: FoodRun 12312345Min: 0.19 / Avg: 0.32 / Max: 0.33Min: 0.32 / Avg: 0.33 / Max: 0.33Min: 0.31 / Avg: 0.31 / Max: 0.32

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: LuxCore BenchmarkRun 1230.1620.3240.4860.6480.81SE +/- 0.02, N = 12SE +/- 0.00, N = 3SE +/- 0.00, N = 30.710.720.71MIN: 0.07 / MAX: 0.86MIN: 0.14 / MAX: 0.86MIN: 0.14 / MAX: 0.85
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: LuxCore BenchmarkRun 123246810Min: 0.49 / Avg: 0.71 / Max: 0.73Min: 0.72 / Avg: 0.72 / Max: 0.73Min: 0.71 / Avg: 0.71 / Max: 0.72

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: Rainbow Colors and PrismRun 1230.7111.4222.1332.8443.555SE +/- 0.06, N = 12SE +/- 0.01, N = 3SE +/- 0.00, N = 33.103.163.09MIN: 0.69 / MAX: 3.3MIN: 2.2 / MAX: 3.3MIN: 2.41 / MAX: 3.21
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender OpenCL 2.3Scene: Rainbow Colors and PrismRun 123246810Min: 2.4 / Avg: 3.1 / Max: 3.19Min: 3.14 / Avg: 3.16 / Max: 3.16Min: 3.09 / Avg: 3.09 / Max: 3.1

FAHBench

FAHBench is a Folding@Home benchmark on the GPU. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterFAHBench 2.3.2Run 1231122334455SE +/- 0.17, N = 3SE +/- 0.16, N = 3SE +/- 0.06, N = 349.4049.7649.06
OpenBenchmarking.orgNs Per Day, More Is BetterFAHBench 2.3.2Run 1231020304050Min: 49.09 / Avg: 49.4 / Max: 49.68Min: 49.49 / Avg: 49.76 / Max: 50.05Min: 48.97 / Avg: 49.06 / Max: 49.17

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 0.26Backend: OpenCLRun 1237001400210028003500SE +/- 50.87, N = 3SE +/- 24.89, N = 3SE +/- 43.33, N = 33444341333371. (CXX) g++ options: -flto -pthread
OpenBenchmarking.orgNodes Per Second, More Is BetterLeelaChessZero 0.26Backend: OpenCLRun 1236001200180024003000Min: 3390 / Avg: 3444.33 / Max: 3546Min: 3375 / Avg: 3413.33 / Max: 3460Min: 3294 / Avg: 3337.33 / Max: 34241. (CXX) g++ options: -flto -pthread

ArrayFire

ArrayFire is an GPU and CPU numeric processing library, this test uses the built-in CPU and OpenCL ArrayFire benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterArrayFire 3.7Test: Conjugate Gradient OpenCLRun 123246810SE +/- 0.004, N = 3SE +/- 0.002, N = 3SE +/- 0.003, N = 38.2968.2978.4201. (CXX) g++ options: -rdynamic
OpenBenchmarking.orgms, Fewer Is BetterArrayFire 3.7Test: Conjugate Gradient OpenCLRun 1233691215Min: 8.29 / Avg: 8.3 / Max: 8.3Min: 8.29 / Avg: 8.3 / Max: 8.3Min: 8.42 / Avg: 8.42 / Max: 8.431. (CXX) g++ options: -rdynamic

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.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: squeezenetRun 1233691215SE +/- 0.02, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 312.4412.5812.91MIN: 11.77 / MAX: 12.86MIN: 11.71 / MAX: 15.31MIN: 11.86 / MAX: 18.761. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: squeezenetRun 12348121620Min: 12.41 / Avg: 12.44 / Max: 12.46Min: 12.56 / Avg: 12.58 / Max: 12.61Min: 12.88 / Avg: 12.91 / Max: 12.951. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mobilenetRun 1233691215SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.02, N = 310.8410.9211.19MIN: 9.36 / MAX: 12.61MIN: 9.24 / MAX: 13.32MIN: 9.62 / MAX: 15.491. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mobilenetRun 1233691215Min: 10.82 / Avg: 10.84 / Max: 10.86Min: 10.92 / Avg: 10.92 / Max: 10.93Min: 11.16 / Avg: 11.19 / Max: 11.221. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2Run 1230.74931.49862.24792.99723.7465SE +/- 0.02, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 33.213.223.33MIN: 3.13 / MAX: 4.86MIN: 3.14 / MAX: 4.81MIN: 3.13 / MAX: 5.71. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2Run 123246810Min: 3.18 / Avg: 3.21 / Max: 3.23Min: 3.22 / Avg: 3.22 / Max: 3.22Min: 3.33 / Avg: 3.33 / Max: 3.331. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3Run 1230.89551.7912.68653.5824.4775SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.03, N = 33.773.783.98MIN: 3.67 / MAX: 5.48MIN: 3.67 / MAX: 5.63MIN: 3.67 / MAX: 7.061. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3Run 123246810Min: 3.74 / Avg: 3.77 / Max: 3.79Min: 3.77 / Avg: 3.78 / Max: 3.78Min: 3.95 / Avg: 3.98 / Max: 4.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: shufflenet-v2Run 1230.631.261.892.523.15SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 32.622.642.80MIN: 2.51 / MAX: 4.09MIN: 2.55 / MAX: 4.15MIN: 2.55 / MAX: 5.671. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: shufflenet-v2Run 123246810Min: 2.61 / Avg: 2.62 / Max: 2.63Min: 2.64 / Avg: 2.64 / Max: 2.64Min: 2.78 / Avg: 2.8 / Max: 2.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mnasnetRun 1230.78081.56162.34243.12323.904SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 33.353.363.47MIN: 3.29 / MAX: 4.84MIN: 3.28 / MAX: 4.97MIN: 3.27 / MAX: 5.371. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: mnasnetRun 123246810Min: 3.33 / Avg: 3.35 / Max: 3.37Min: 3.35 / Avg: 3.36 / Max: 3.37Min: 3.46 / Avg: 3.47 / Max: 3.491. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: efficientnet-b0Run 123246810SE +/- 0.03, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 35.785.846.06MIN: 5.55 / MAX: 9.45MIN: 5.6 / MAX: 8.14MIN: 5.63 / MAX: 9.361. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: efficientnet-b0Run 123246810Min: 5.73 / Avg: 5.78 / Max: 5.82Min: 5.79 / Avg: 5.84 / Max: 5.95Min: 6.05 / Avg: 6.06 / Max: 6.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: blazefaceRun 1230.24750.4950.74250.991.2375SE +/- 0.01, N = 3SE +/- 0.00, N = 3SE +/- 0.01, N = 31.011.011.10MIN: 1 / MAX: 1.23MAX: 1.16MIN: 1.01 / MAX: 3.241. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: blazefaceRun 123246810Min: 1 / Avg: 1.01 / Max: 1.02Min: 1.01 / Avg: 1.01 / Max: 1.02Min: 1.08 / Avg: 1.1 / Max: 1.121. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: googlenetRun 123246810SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 38.578.568.90MIN: 8.28 / MAX: 11.09MIN: 8.39 / MAX: 9.07MIN: 8.49 / MAX: 10.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: googlenetRun 1233691215Min: 8.56 / Avg: 8.57 / Max: 8.6Min: 8.55 / Avg: 8.56 / Max: 8.57Min: 8.89 / Avg: 8.9 / Max: 8.921. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: vgg16Run 123816243240SE +/- 0.01, N = 3SE +/- 0.03, N = 3SE +/- 0.03, N = 334.0035.2136.20MIN: 33.54 / MAX: 36.65MIN: 33.68 / MAX: 38.75MIN: 34.35 / MAX: 41.961. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: vgg16Run 123816243240Min: 33.99 / Avg: 34 / Max: 34.02Min: 35.14 / Avg: 35.21 / Max: 35.25Min: 36.16 / Avg: 36.2 / Max: 36.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet18Run 123246810SE +/- 0.00, N = 3SE +/- 0.02, N = 3SE +/- 0.01, N = 36.246.376.60MIN: 6.11 / MAX: 8.26MIN: 6.02 / MAX: 8.14MIN: 6.16 / MAX: 10.351. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet18Run 1233691215Min: 6.23 / Avg: 6.24 / Max: 6.24Min: 6.33 / Avg: 6.37 / Max: 6.41Min: 6.57 / Avg: 6.6 / Max: 6.611. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: alexnetRun 1233691215SE +/- 0.01, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 38.818.869.01MIN: 8.56 / MAX: 11.16MIN: 8.62 / MAX: 10.61MIN: 8.66 / MAX: 12.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: alexnetRun 1233691215Min: 8.8 / Avg: 8.81 / Max: 8.82Min: 8.84 / Avg: 8.86 / Max: 8.87Min: 8.97 / Avg: 9.01 / Max: 9.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet50Run 12348121620SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 314.3415.1115.75MIN: 13.73 / MAX: 16.79MIN: 13.88 / MAX: 17.23MIN: 14.11 / MAX: 20.921. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: resnet50Run 12348121620Min: 14.27 / Avg: 14.34 / Max: 14.39Min: 15.06 / Avg: 15.11 / Max: 15.14Min: 15.7 / Avg: 15.75 / Max: 15.841. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: yolov4-tinyRun 123510152025SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.06, N = 320.0520.1020.04MIN: 15.64 / MAX: 21.87MIN: 18.89 / MAX: 21.77MIN: 17.33 / MAX: 29.731. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20200916Target: Vulkan GPU - Model: yolov4-tinyRun 123510152025Min: 20.02 / Avg: 20.05 / Max: 20.08Min: 20.06 / Avg: 20.1 / Max: 20.15Min: 19.98 / Avg: 20.04 / Max: 20.161. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

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 2.90Blend File: BMW27 - Compute: CUDARun 12380160240320400SE +/- 1.08, N = 3SE +/- 0.44, N = 3SE +/- 0.54, N = 3308.30381.11381.65
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: CUDARun 12370140210280350Min: 306.44 / Avg: 308.3 / Max: 310.19Min: 380.24 / Avg: 381.11 / Max: 381.63Min: 380.7 / Avg: 381.65 / Max: 382.57

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CUDARun 1232004006008001000SE +/- 1.20, N = 3SE +/- 0.89, N = 3SE +/- 0.60, N = 3894.18912.44912.76
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: CUDARun 123160320480640800Min: 891.79 / Avg: 894.18 / Max: 895.59Min: 910.67 / Avg: 912.44 / Max: 913.34Min: 911.63 / Avg: 912.76 / Max: 913.65

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CUDARun 1232004006008001000SE +/- 0.04, N = 3SE +/- 0.15, N = 3SE +/- 0.31, N = 3646.84827.47827.39
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: CUDARun 123150300450600750Min: 646.77 / Avg: 646.84 / Max: 646.9Min: 827.19 / Avg: 827.47 / Max: 827.72Min: 826.78 / Avg: 827.39 / Max: 827.8

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CUDARun 1237001400210028003500SE +/- 33.13, N = 9SE +/- 0.54, N = 3SE +/- 0.19, N = 33167.763297.833297.57
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Barbershop - Compute: CUDARun 1236001200180024003000Min: 3100.71 / Avg: 3167.76 / Max: 3321.62Min: 3297.24 / Avg: 3297.83 / Max: 3298.91Min: 3297.2 / Avg: 3297.57 / Max: 3297.76

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: NVIDIA OptiXRun 12390180270360450SE +/- 7.22, N = 9SE +/- 0.19, N = 3SE +/- 0.14, N = 3349.29403.46403.51
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: BMW27 - Compute: NVIDIA OptiXRun 12370140210280350Min: 341.62 / Avg: 349.29 / Max: 407.05Min: 403.18 / Avg: 403.46 / Max: 403.81Min: 403.23 / Avg: 403.51 / Max: 403.68

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: NVIDIA OptiXRun 1232004006008001000SE +/- 2.26, N = 3SE +/- 0.58, N = 3SE +/- 0.08, N = 31021.551099.241100.13
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Classroom - Compute: NVIDIA OptiXRun 1232004006008001000Min: 1017.03 / Avg: 1021.55 / Max: 1023.93Min: 1098.09 / Avg: 1099.24 / Max: 1099.87Min: 1099.97 / Avg: 1100.13 / Max: 1100.25

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: NVIDIA OptiXRun 1232004006008001000SE +/- 0.10, N = 3SE +/- 1.20, N = 3SE +/- 0.39, N = 3715.41949.11950.93
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Fishy Cat - Compute: NVIDIA OptiXRun 123170340510680850Min: 715.24 / Avg: 715.41 / Max: 715.57Min: 946.9 / Avg: 949.11 / Max: 951.04Min: 950.38 / Avg: 950.93 / Max: 951.68

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CUDARun 1235001000150020002500SE +/- 5.87, N = 3SE +/- 2.93, N = 3SE +/- 2.34, N = 31896.872114.702109.62
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: CUDARun 123400800120016002000Min: 1885.14 / Avg: 1896.87 / Max: 1903.18Min: 2108.84 / Avg: 2114.7 / Max: 2117.88Min: 2104.95 / Avg: 2109.62 / Max: 2112.16

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXRun 1235001000150020002500SE +/- 0.08, N = 3SE +/- 0.17, N = 3SE +/- 0.19, N = 31960.712301.382301.04
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.90Blend File: Pabellon Barcelona - Compute: NVIDIA OptiXRun 123400800120016002000Min: 1960.56 / Avg: 1960.71 / Max: 1960.82Min: 2301.18 / Avg: 2301.38 / Max: 2301.72Min: 2300.71 / Avg: 2301.04 / Max: 2301.37

MandelGPU

MandelGPU is an OpenCL benchmark and this test runs with the OpenCL rendering float4 kernel with a maximum of 4096 iterations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSamples/sec, More Is BetterMandelGPU 1.3pts1OpenCL Device: GPURun 12312M24M36M48M60MSE +/- 69191.69, N = 3SE +/- 291084.10, N = 3SE +/- 364793.78, N = 354591952.553723378.254824518.71. (CC) gcc options: -O3 -lm -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL
OpenBenchmarking.orgSamples/sec, More Is BetterMandelGPU 1.3pts1OpenCL Device: GPURun 12310M20M30M40M50MMin: 54465918.4 / Avg: 54591952.47 / Max: 54704457.2Min: 53253601.8 / Avg: 53723378.2 / Max: 54256045.6Min: 54379141.9 / Avg: 54824518.67 / Max: 555476601. (CC) gcc options: -O3 -lm -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL

clpeak

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

OpenBenchmarking.orgGIOPS, More Is BetterclpeakOpenCL Test: Integer Compute INTRun 123120240360480600SE +/- 6.15, N = 3SE +/- 6.20, N = 3SE +/- 4.47, N = 3542.12531.80532.001. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGIOPS, More Is BetterclpeakOpenCL Test: Integer Compute INTRun 123100200300400500Min: 530.77 / Avg: 542.12 / Max: 551.88Min: 524.34 / Avg: 531.8 / Max: 544.11Min: 524.33 / Avg: 532 / Max: 539.81. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatRun 123400800120016002000SE +/- 31.67, N = 15SE +/- 25.82, N = 15SE +/- 33.84, N = 151870.181859.871842.121. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatRun 12330060090012001500Min: 1600.93 / Avg: 1870.18 / Max: 1932.41Min: 1659.69 / Avg: 1859.87 / Max: 1912.05Min: 1576.88 / Avg: 1842.12 / Max: 1921.931. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleRun 1231530456075SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.15, N = 366.6665.6265.831. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleRun 1231326395265Min: 66.57 / Avg: 66.66 / Max: 66.72Min: 65.53 / Avg: 65.62 / Max: 65.76Min: 65.63 / Avg: 65.83 / Max: 66.131. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthRun 12320406080100SE +/- 0.03, N = 3SE +/- 0.07, N = 3SE +/- 0.06, N = 390.6889.3289.591. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthRun 12320406080100Min: 90.62 / Avg: 90.68 / Max: 90.73Min: 89.23 / Avg: 89.32 / Max: 89.46Min: 89.49 / Avg: 89.59 / Max: 89.691. (CXX) g++ options: -O3 -rdynamic -lOpenCL

NeatBench

NeatBench is a benchmark of the cross-platform Neat Video software on the CPU and optional GPU (OpenCL / CUDA) support. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterNeatBench 5Acceleration: GPURun 1233691215SE +/- 0.00, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 39.529.399.38
OpenBenchmarking.orgFPS, More Is BetterNeatBench 5Acceleration: GPURun 1233691215Min: 9.52 / Avg: 9.52 / Max: 9.52Min: 9.38 / Avg: 9.39 / Max: 9.4Min: 9.36 / Avg: 9.38 / Max: 9.4