To reduce the scope of the current PR, I am going to undo the change from dense-to-sparse storage of the diagonal Hodge star on CUDA. Some benchmarks need to be made to test this change and that is best handled outside of this refactor.
Originally posted by @lukem12345 in #227 (comment)
To reduce the scope of the current PR, I am going to undo the change from dense-to-sparse storage of the diagonal Hodge star on CUDA. Some benchmarks need to be made to test this change and that is best handled outside of this refactor.
Originally posted by @lukem12345 in #227 (comment)