Newer
Older
On a desktop system this is probably not required. On a cluster this is usually required to make sure that an appropriate
- module load gcc (tested on gcc 4.9)
It is important to be sure that your CUDA environment is properly setup too (nvcc must be included on PATH).
- PATH=$PATH:[CUDA_HOME]/bin
- make -j
- unit_test1, cuThomasBatch execution
- unit_test2, execution of cuThomasBatch, cuSparse and cuSolve reference routines
- ./unit_test[1,2] [SIZE_SYSTEM] [BATCH_COUNT] [CUDA_BLOCK_SIZE]