Add CUDA support for the DSL
Created by: maedoc
From @DeLaVlag and @sdiazpier, developed in the TVB-HPC repository, extends DSL for generating full simulation loops for GPU-based parameter sweeps with the DSL.
@DeLaVlag if you want to push some commits here, please fork this repo, checkout this branch, push the branch to your fork and we can redo the PR from your copy.
small checklist prior to merge
-
get a review -
notebook to demo running one of the generated kernels