Enable register spilling to shared memory #1132
Draft
Conversation
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
CUDA 13.0 enables the PTX assembler to spill registers to shared memory instead of local memory, which should both be much faster, and also reduce the local memory usage of our fitting and finding kernels which are currently bottlenecking our throughput.
|
|
I'm not 100% certain this works as intended like this, as this pragma is to be attached at the function scope. But we can try. |
|
This is interesting as we are not actively using the shared memory in our finding and fitting kernels. |
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Outdated suggestions cannot be applied.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.
CUDA 13.0 enables the PTX assembler to spill registers to shared memory instead of local memory, which should both be much faster, and also reduce the local memory usage of our fitting and finding kernels which are currently bottlenecking our throughput.