Update spmm naive kernel for warpsize 64 #50
Merged
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
|
@pnunna93 Did you run the unit tests? |
lcskrishna
requested changes
Nov 8, 2024
csrc/kernels.hip
Outdated
| @@ -2853,6 +2853,7 @@ template <int THREADS, int ITEMS_PER_THREAD, int TILE_ROWS, int TILE_COLS, int T | |||
| #define DENORM 1.0f/127.0f | |||
| #define MAX_SPARSE_COUNT 32 | |||
| #define SMEM_SIZE 8*256 | |||
| #define WARP_SIZE __AMDGCN_WAVEFRONT_SIZE | |||
There was a problem hiding this comment.
Instead of __AMDGCN_WAVEFRONT_SIZE can we use the warpSize defined in hip/hip_runtime.h header file.
#define WARP_SIZE warpSize
For example: https://github.com/pytorch/pytorch/blob/2cefbb71cf65696e28ee1bdfce06d6846285e5fb/c10/macros/Macros.h#L322
There was a problem hiding this comment.
Changed it. I have modified 4 bit gemm kernel also to use the same.
Yes. This is latest log - BnB_UT_Summary_1108.log |
lcskrishna
approved these changes
Nov 11, 2024
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Outdated suggestions cannot be applied.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 adds a macro to set warpsize based on gpu for spmm naive kernel