This sample application performs general matrix multiplication using OpenCL(TM) GPU device and provides some statistics based on manual source code instrumentation with OpenCL built-in intrinsics.
OpenCL Matrix Multiplication (matrix size: 1024 x 1024, repeats 4 times)
Target device: Intel(R) Gen9 HD Graphics NEO
Matrix multiplication time: 0.073987 sec
Results are CORRECT with accuracy: 4.90573e-06
Matrix multiplication time: 0.069171 sec
Results are CORRECT with accuracy: 4.90573e-06
Matrix multiplication time: 0.0688272 sec
Results are CORRECT with accuracy: 4.90573e-06
Matrix multiplication time: 0.0686009 sec
Results are CORRECT with accuracy: 4.90573e-06
Total execution time: 0.56824 sec
======== GEMM Instrumentation Results ========
Samples collected: 131072 (16.54% of bad samples)
Estimated average HW thread duration: ~ 454 us (454972 cycles)
Estimated number of HW threads per EU: 7
Estimated number of EUs per subslice: 8
Estimated number of subslices per dual subslice: 3
Estimated number of dual subslices per slice: 1
Estimated number of slices: 2
Estimated total number EUs: 48
Estimated total number of HW threads: 336
==============================================
- Linux
- Windows
- CMake (version 3.12 and above)
- Git (version 1.8 and above)
- Python (version 2.7 and above)
- OpenCL(TM) ICD Loader
- Intel(R) Graphics Compute Runtime for oneAPI Level Zero and OpenCL(TM) Driver
Run the following commands to build the sample:
cd <pti>/samples/cl_gemm_inst
mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=Release ..
makeUse this command line to run the application:
./bin/cl_gemm_inst [matrix_size] [repeat_count]Use Microsoft* Visual Studio x64 command prompt to run the following commands and build the sample:
cd <pti>\samples\cl_gemm_inst
mkdir build
cd build
cmake -G "NMake Makefiles" -DCMAKE_BUILD_TYPE=Release -DCMAKE_LIBRARY_PATH=<opencl_icd_lib_path> ..
nmakeUse this command line to run the application:
cl_gemm_inst.exe [matrix_size] [repeats_count]