Reuse GPU E-vecs #1648
Merged
Reuse GPU E-vecs #1648
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jeremylt
force-pushed
the
jeremy/cuda-reuse-out
branch
3 times, most recently
from
95c081e to
bcaca9e
Compare
|
I did not add this change on the CPU side, as a 1 or 8 element block for the Opt backends is pretty negligible memory that doesn't seem worth the extra complexity in the code compared to the memory savings on the GPU side |
jrwrigh
approved these changes
Aug 29, 2024
jeremylt
force-pushed
the
jeremy/cuda-reuse-out
branch
from
bcaca9e to
8a21357
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR reuses active input e-vec buffers for output e-vecs where able. These are one of the biggest allocations of memory that the user doesn't see.
Adding this for AtPoints is a bit trickier. It is probably best as a follow-upNote: manually tested with Ratel as well