Skip to content

Solver SVD Optimizations and Improved cuSolver batching#721

Merged
cliffburdick merged 3 commits intomainfrom
solver_updates
Aug 19, 2024
Merged

Solver SVD Optimizations and Improved cuSolver batching#721
cliffburdick merged 3 commits intomainfrom
solver_updates

Conversation

@aayushg55
Copy link
Contributor

  • Bypassed temp allocations and having to transpose outputs in most cases for SVD
  • Added m < n CUDA SVD support (cuSolver natively only supports m >= n)
  • Added option to use gesdd or gesvd LAPACK driver for Host SVD
  • For cuSolver batching, replaced per-iteration sync & error check with a single check at the end to reduce blocking operations

@aayushg55 aayushg55 marked this pull request as ready for review August 16, 2024 22:08
@aayushg55 aayushg55 requested a review from cliffburdick August 16, 2024 22:09
@cliffburdick
Copy link
Collaborator

/build

@cliffburdick
Copy link
Collaborator

/build

@cliffburdick
Copy link
Collaborator

/build

1 similar comment
@cliffburdick
Copy link
Collaborator

/build

@cliffburdick cliffburdick merged commit f55bcd3 into main Aug 19, 2024
@cliffburdick cliffburdick deleted the solver_updates branch August 19, 2024 16:05
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

Successfully merging this pull request may close these issues.

2 participants