Due 2024-10-29.

The goal of this homework is to get you up and running CUDA code on Perlmutter. Using code at https://github.com/cs5220-f24/hw3, you should

  • Edit vec_add.cu to actually use CUDA with a grid-stride loop
  • Adapt the submission script to run with nvprof on Perlmutter
  • Submit the vec_add.cu code and the output of nvprof

See the class slides and https://developer.nvidia.com/blog/even-easier-introduction-cuda/