Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improving performance on GPUs (2) - reduce getindex() and setindex!() time #3

Closed
youngdae opened this issue Dec 4, 2020 · 1 comment

Comments

@youngdae
Copy link
Member

youngdae commented Dec 4, 2020

In the incomplete Cholesky factorization routine, a significant amount of time is being used in getindex() and setindex!() routine. This seems related to computing the entries of L. We need to investigate a more GPU friendly way of doing this.

Candidate algorithms:

  • We could just use NVIDIA's incomplete Cholesky factorization routine (with zero fill-in), but its performance could be very different from ours, because their implementation is quite different from ours.
  • We could investigate a GPU friendly way of performing computation, especially focusing on small matrices.
@kibaekkim
Copy link
Member

This looks relevant to the first candidate: http://juliasmoothoptimizers.github.io/Krylov.jl/dev/gpu/#Example-with-a-general-square-system Once I get more familiarized to this, I can try that.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants