OpenCL ROCm 2.0 vs. AMDGPU-PRO Linux

Radeon RX Vega 64 ROCm 2.0 OpenCL versus PAL OpenCL driver in AMDGPU-PRO 18.50. Benchmarks by Michael Larabel for a future article on Phoronix.com.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 1901167-PTS-OPENCLRO89
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
HPC - High Performance Computing 4 Tests
Machine Learning 3 Tests
NVIDIA GPU Compute 6 Tests
OpenCL 6 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
ROCm 2.0
January 16 2019
  55 Minutes
AMDGPU-PRO 18.50 PAL
January 16 2019
  23 Minutes
Invert Hiding All Results Option
  39 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):


OpenCL ROCm 2.0 vs. AMDGPU-PRO LinuxOpenBenchmarking.orgPhoronix Test Suite 10.6.1AMD Ryzen Threadripper 2990WX 32-Core @ 3.00GHz (32 Cores / 64 Threads)ASUS ROG ZENITH EXTREME (1601 BIOS)AMD Family 17h32768MB16GB Voyager 3.0 + Samsung SSD 970 EVO 500GBAMD Radeon RX Vega 8GB (1630/945MHz)Realtek ALC1220ASUS VP28UIntel I211 + Qualcomm Atheros QCA6174 802.11ac + Wilocity Wil6200 802.11adUbuntu 18.044.15.0-43-generic (x86_64)GNOME Shell 3.28.3X Server 1.19.6amdgpu 18.0.1amdgpu 18.1.994.5 Mesa 18.0.5 (LLVM 6.0.0)4.6.13542GCC 7.3.0ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriversOpenGLsCompilerFile-SystemScreen ResolutionOpenCL ROCm 2.0 Vs. AMDGPU-PRO Linux BenchmarksSystem Logs- --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++ --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: acpi-cpufreq ondemand- GLAMOR- Python 2.7.15rc1 + Python 3.6.7- __user pointer sanitization + Full AMD retpoline IBPB + SSB disabled via prctl and seccomp

ROCm 2.0 vs. AMDGPU-PRO 18.50 PAL ComparisonPhoronix Test Suite 10.6.1Baseline+73.9%+73.9%+147.8%+147.8%+221.7%+221.7%148.8%76.9%64.7%44%19.8%7.1%6.1%4.4%4.2%3.6%Kernel Latency295.4%ReadT.B.e102%Boat - OpenCL89.7%Server Rack - OpenCLCopyT.B.e56.2%GPUServer Room - OpenCL30.5%OpenCL - FFT SP24.6%No - Inference - Inception V3 - OpenCLOpenCL - Triad10.2%OpenCL - MD5 HashOpenCL HeartwallNo - Inference - IMDB LSTM - OpenCLMasskrug - OpenCLS.P.F4.2%OpenCL - T.R.B4%No - Inference - ResNet 50 - OpenCLclpeakcl-memclpeakDarktableDarktablecl-memclpeakJuliaGPUDarktableSHOC Scalable HeterOgeneous ComputingPlaidMLSHOC Scalable HeterOgeneous ComputingSHOC Scalable HeterOgeneous ComputingRodiniaPlaidMLDarktableclpeakSHOC Scalable HeterOgeneous ComputingPlaidMLROCm 2.0AMDGPU-PRO 18.50 PAL

OpenCL ROCm 2.0 vs. AMDGPU-PRO Linuxshoc: OpenCL - Triadshoc: OpenCL - FFT SPshoc: OpenCL - MD5 Hashshoc: OpenCL - Bus Speed Downloadshoc: OpenCL - Bus Speed Readbackshoc: OpenCL - Texture Read Bandwidthcl-mem: Copycl-mem: Readcl-mem: Writeplaidml: No - Inference - IMDB LSTM - OpenCLplaidml: No - Inference - Mobilenet - OpenCLplaidml: No - Inference - ResNet 50 - OpenCLplaidml: No - Inference - Inception V3 - OpenCLlczero: OpenCLrodinia: OpenCL Heartwalldarktable: Boat - OpenCLdarktable: Masskrug - OpenCLdarktable: Server Rack - OpenCLdarktable: Server Room - OpenCLjuliagpu: GPUclpeak: Kernel Latencyclpeak: Integer Compute INTclpeak: Single-Precision Floatclpeak: Double-Precision Doubleclpeak: Global Memory Bandwidthclpeak: Transfer Bandwidth enqueueReadBufferclpeak: Transfer Bandwidth enqueueWriteBufferROCm 2.0AMDGPU-PRO 18.50 PAL6.69107516.467.147.16441221160384252479225113.883044.034.485.700.231.9716685881610.5624971305383336217.0645.446.0786317.637.147.16424364398379263479233136.383.808.505.470.132.5724029346941.7524861252882836110.9222.49OpenBenchmarking.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: TriadROCm 2.0AMDGPU-PRO 18.50 PAL246810SE +/- 0.01, N = 3SE +/- 0.02, N = 36.696.071. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: TriadROCm 2.0AMDGPU-PRO 18.50 PAL3691215Min: 6.67 / Avg: 6.69 / Max: 6.7Min: 6.05 / Avg: 6.07 / Max: 6.111. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: FFT SPROCm 2.0AMDGPU-PRO 18.50 PAL2004006008001000SE +/- 1.63, N = 3SE +/- 11.99, N = 610758631. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGFLOPS, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: FFT SPROCm 2.0AMDGPU-PRO 18.50 PAL2004006008001000Min: 1071.9 / Avg: 1074.66 / Max: 1077.55Min: 803.24 / Avg: 862.6 / Max: 881.711. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: MD5 HashROCm 2.0AMDGPU-PRO 18.50 PAL48121620SE +/- 0.05, N = 3SE +/- 0.03, N = 316.4617.631. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGHash/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: MD5 HashROCm 2.0AMDGPU-PRO 18.50 PAL48121620Min: 16.37 / Avg: 16.46 / Max: 16.54Min: 17.59 / Avg: 17.63 / Max: 17.691. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed DownloadROCm 2.0AMDGPU-PRO 18.50 PAL246810SE +/- 0.00, N = 3SE +/- 0.00, N = 37.147.141. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed DownloadROCm 2.0AMDGPU-PRO 18.50 PAL3691215Min: 7.14 / Avg: 7.14 / Max: 7.15Min: 7.14 / Avg: 7.14 / Max: 7.151. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed ReadbackROCm 2.0AMDGPU-PRO 18.50 PAL246810SE +/- 0.00, N = 3SE +/- 0.00, N = 37.167.161. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Bus Speed ReadbackROCm 2.0AMDGPU-PRO 18.50 PAL3691215Min: 7.15 / Avg: 7.16 / Max: 7.16Min: 7.16 / Avg: 7.16 / Max: 7.161. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Texture Read BandwidthROCm 2.0AMDGPU-PRO 18.50 PAL100200300400500SE +/- 1.31, N = 3SE +/- 1.67, N = 34414241. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt
OpenBenchmarking.orgGB/s, More Is BetterSHOC Scalable HeterOgeneous Computing 2015-11-10Target: OpenCL - Benchmark: Texture Read BandwidthROCm 2.0AMDGPU-PRO 18.50 PAL80160240320400Min: 439.42 / Avg: 441.32 / Max: 443.82Min: 420.55 / Avg: 423.89 / Max: 425.671. (CXX) g++ options: -O2 -lSHOCCommonOpenCL -lSHOCCommon -lOpenCL -lrt

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: CopyROCm 2.0AMDGPU-PRO 18.50 PAL80160240320400SE +/- 0.06, N = 3SE +/- 0.00, N = 32213641. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: CopyROCm 2.0AMDGPU-PRO 18.50 PAL70140210280350Min: 221.3 / Avg: 221.4 / Max: 221.5Min: 364.1 / Avg: 364.1 / Max: 364.11. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadROCm 2.0AMDGPU-PRO 18.50 PAL90180270360450SE +/- 0.06, N = 3SE +/- 0.15, N = 31603981. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: ReadROCm 2.0AMDGPU-PRO 18.50 PAL70140210280350Min: 159.6 / Avg: 159.7 / Max: 159.8Min: 397.2 / Avg: 397.5 / Max: 397.71. (CC) gcc options: -O2 -flto -lOpenCL

OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteROCm 2.0AMDGPU-PRO 18.50 PAL80160240320400SE +/- 0.07, N = 3SE +/- 0.82, N = 33843791. (CC) gcc options: -O2 -flto -lOpenCL
OpenBenchmarking.orgGB/s, More Is Bettercl-mem 2017-01-13Benchmark: WriteROCm 2.0AMDGPU-PRO 18.50 PAL70140210280350Min: 384.3 / Avg: 384.37 / Max: 384.5Min: 377.7 / Avg: 379.33 / Max: 380.21. (CC) gcc options: -O2 -flto -lOpenCL

PlaidML

This test profile uses PlaidML deep learning framework for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: IMDB LSTM - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL60120180240300SE +/- 0.35, N = 3SE +/- 0.88, N = 3252263
OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: IMDB LSTM - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL50100150200250Min: 251.26 / Avg: 251.71 / Max: 252.4Min: 261.34 / Avg: 262.98 / Max: 264.35

OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Mobilenet - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL100200300400500SE +/- 0.09, N = 3SE +/- 0.67, N = 3479479
OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Mobilenet - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL80160240320400Min: 479.12 / Avg: 479.29 / Max: 479.42Min: 477.7 / Avg: 478.54 / Max: 479.85

OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL50100150200250SE +/- 0.86, N = 3SE +/- 1.16, N = 3225233
OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL4080120160200Min: 222.8 / Avg: 224.52 / Max: 225.51Min: 231.76 / Avg: 233.49 / Max: 235.68

OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Inception V3 - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL306090120150SE +/- 1.83, N = 12SE +/- 0.61, N = 3113.88136.38
OpenBenchmarking.orgExamples Per Second, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: Inception V3 - Device: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL306090120150Min: 106.83 / Avg: 113.88 / Max: 120Min: 135.16 / Avg: 136.38 / Max: 137.14

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: OpenCLROCm 2.070140210280350SE +/- 1.46, N = 33041. (CXX) g++ options: -lpthread -lz

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes the OpenCL and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenCL HeartwallROCm 2.0AMDGPU-PRO 18.50 PAL0.90681.81362.72043.62724.534SE +/- 0.00, N = 3SE +/- 0.01, N = 34.033.801. (CXX) g++ options: -O2 -lOpenCL
OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenCL HeartwallROCm 2.0AMDGPU-PRO 18.50 PAL246810Min: 4.02 / Avg: 4.03 / Max: 4.03Min: 3.79 / Avg: 3.8 / Max: 3.811. (CXX) g++ options: -O2 -lOpenCL

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 2.4.2Test: Boat - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL246810SE +/- 0.08, N = 12SE +/- 0.08, N = 34.488.50
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Boat - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL3691215Min: 3.95 / Avg: 4.48 / Max: 4.88Min: 8.4 / Avg: 8.5 / Max: 8.65

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Masskrug - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL1.28252.5653.84755.136.4125SE +/- 0.02, N = 3SE +/- 0.04, N = 35.705.47
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Masskrug - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL246810Min: 5.67 / Avg: 5.7 / Max: 5.73Min: 5.41 / Avg: 5.47 / Max: 5.56

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Server Rack - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL0.05180.10360.15540.20720.259SE +/- 0.00, N = 3SE +/- 0.00, N = 30.230.13
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Server Rack - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL12345Min: 0.23 / Avg: 0.23 / Max: 0.23Min: 0.13 / Avg: 0.13 / Max: 0.14

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Server Room - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL0.57831.15661.73492.31322.8915SE +/- 0.01, N = 3SE +/- 0.04, N = 41.972.57
OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.4.2Test: Server Room - Acceleration: OpenCLROCm 2.0AMDGPU-PRO 18.50 PAL246810Min: 1.96 / Avg: 1.97 / Max: 1.98Min: 2.45 / Avg: 2.57 / Max: 2.64

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: GPUROCm 2.0AMDGPU-PRO 18.50 PAL50M100M150M200M250MSE +/- 1617179.19, N = 3SE +/- 1514546.33, N = 31668588162402934691. (CC) gcc options: -O3 -march=native -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL -lm
OpenBenchmarking.orgSamples/sec, More Is BetterJuliaGPU 1.2pts1OpenCL Device: GPUROCm 2.0AMDGPU-PRO 18.50 PAL40M80M120M160M200MMin: 165235253.8 / Avg: 166858815.63 / Max: 170093165.6Min: 238388090.3 / Avg: 240293468.7 / Max: 243285447.41. (CC) gcc options: -O3 -march=native -ftree-vectorize -funroll-loops -lglut -lOpenCL -lGL -lm

clpeak

OpenBenchmarking.orgus, Fewer Is BetterclpeakOpenCL Test: Kernel LatencyROCm 2.0AMDGPU-PRO 18.50 PAL1020304050SE +/- 0.14, N = 3SE +/- 1.09, N = 1210.5641.751. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgus, Fewer Is BetterclpeakOpenCL Test: Kernel LatencyROCm 2.0AMDGPU-PRO 18.50 PAL918273645Min: 10.31 / Avg: 10.56 / Max: 10.81Min: 31.69 / Avg: 41.75 / Max: 45.61. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGIOPS, More Is BetterclpeakOpenCL Test: Integer Compute INTROCm 2.0AMDGPU-PRO 18.50 PAL5001000150020002500SE +/- 2.51, N = 3SE +/- 6.75, N = 3249724861. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGIOPS, More Is BetterclpeakOpenCL Test: Integer Compute INTROCm 2.0AMDGPU-PRO 18.50 PAL400800120016002000Min: 2493.63 / Avg: 2497 / Max: 2501.91Min: 2474.49 / Avg: 2485.54 / Max: 2497.781. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatROCm 2.0AMDGPU-PRO 18.50 PAL3K6K9K12K15KSE +/- 11.58, N = 3SE +/- 40.57, N = 313053125281. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Single-Precision FloatROCm 2.0AMDGPU-PRO 18.50 PAL2K4K6K8K10KMin: 13031.22 / Avg: 13053.31 / Max: 13070.38Min: 12447.49 / Avg: 12528.44 / Max: 12573.781. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleROCm 2.0AMDGPU-PRO 18.50 PAL2004006008001000SE +/- 0.95, N = 3SE +/- 1.70, N = 38338281. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGFLOPS, More Is BetterclpeakOpenCL Test: Double-Precision DoubleROCm 2.0AMDGPU-PRO 18.50 PAL150300450600750Min: 831.38 / Avg: 832.58 / Max: 834.45Min: 825.82 / Avg: 827.64 / Max: 831.031. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthROCm 2.0AMDGPU-PRO 18.50 PAL80160240320400SE +/- 0.12, N = 3SE +/- 0.10, N = 33623611. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Global Memory BandwidthROCm 2.0AMDGPU-PRO 18.50 PAL60120180240300Min: 361.94 / Avg: 362.17 / Max: 362.33Min: 360.51 / Avg: 360.68 / Max: 360.871. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueReadBufferROCm 2.0AMDGPU-PRO 18.50 PAL48121620SE +/- 0.03, N = 3SE +/- 0.88, N = 917.0610.921. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueReadBufferROCm 2.0AMDGPU-PRO 18.50 PAL48121620Min: 17.01 / Avg: 17.06 / Max: 17.09Min: 4.7 / Avg: 10.92 / Max: 14.351. (CXX) g++ options: -O3 -rdynamic -lOpenCL

OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueWriteBufferROCm 2.0AMDGPU-PRO 18.50 PAL1020304050SE +/- 0.01, N = 3SE +/- 0.21, N = 345.4422.491. (CXX) g++ options: -O3 -rdynamic -lOpenCL
OpenBenchmarking.orgGBPS, More Is BetterclpeakOpenCL Test: Transfer Bandwidth enqueueWriteBufferROCm 2.0AMDGPU-PRO 18.50 PAL918273645Min: 45.43 / Avg: 45.44 / Max: 45.45Min: 22.08 / Avg: 22.49 / Max: 22.71. (CXX) g++ options: -O3 -rdynamic -lOpenCL