Improve ClBlast implementation, avoid recreating buffers, remove redundant transfers#45
Merged
LostRuins merged 2 commits intoLostRuins:concedofrom Apr 13, 2023
Merged
Conversation
Owner
|
|
Author
|
Keeping the buffers open didn't improve performance at all, so I scrapped that. Buffering matrices in advance is difficult because they are stored in a quantized format and would either have to be stored unquantized on the GPU, which needs a lot of VRAM, or dequantized on the GPU, which would need a new kernel and quite a bit of work. So this PR just does some code improvements and platform + device printing, but performance stays the same. |
Owner
|
LGTM, merged |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Didn't manage to set up caching yet. I improved some steps, but the difference wasn't big.