Setting Up Programming Environment¶

1. Load the compiler module¶

We need the Nvidia HPC SDK compiler for compiling and testing OpenACC code.

$ module load OpenMPI/5.0.3-NVHPC-24.9-CUDA-12.6.0
### or
$ source module.sh

Check if the module is loaded properly

$ ml

Currently Loaded Modules:
  1) NVHPC/24.9-CUDA-12.6.0
  2) CUDA/12.6.0
  3) OpenMPI/5.0.3-NVHPC-24.9-CUDA-12.6.0
  ...

$ nvc -fast -acc=gpu -gpu=cc80 -Minfo=accel <source_file>.c -o <output>

For example:

$ nvc -fast -acc=gpu -gpu=cc80 -Minfo=accel Hello-world.c -o Hello-World-GPU

$ ./<output>

For example:

$ ./Hello-World-GPU

Hello World from GPU!
Hello World from GPU!
Hello World from GPU!
Hello World from GPU!