Using GPU compilers

To compile CUDA code to run on the Geyser and Caldera clusters, use either NVIDIA's nvcc compiler command for CUDA C and C++ code, or use PGI for CUDA Fortran code.

NVIDIA

Load the cuda module, then compile your program.

module load cuda
nvcc foo.cu

Non-GPU compilers (Intel, PGI, GNU) can be loaded at the same time.

PGI

Load the PGI CUDA Fortran compiler module, then compile your program.

module load pgi/cuda-15.1
pgf90 -Mcuda foo.f90

or

pgf90 foo.cuf

If your Fortran program file name includes the .cuf extension, the -Mcuda command line option is not necessary.

For detailed compilation examples, see: