Support different dtype for LinearCombination #556
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
Fix for supporting different data type for LinearCombination Output and Compute.
Antonyvance
added
bug
joyalbin
changed the title
taozha2
reviewed
Oct 23, 2025
| typename TiledMMAHelper<MMA_Atom<XE_8x16x16_F32F16F16F32_TT>, Layout<Shape<_32, _128, _32>>, | ||
| Layout<Shape<_1, _4, _1>, Stride<_4, _1, _0>>>::TiledMMA, | ||
| XE_2D_U16x32x32_LD_N, XE_2D_U4x32x16_LD_T, XE_2D_U16x8x16_ST_N, | ||
| XE_2D_U16x32x32_LD_N, XE_2D_U4x32x16_LD_T, XE_2D_U32x8x16_ST_N, |
There was a problem hiding this comment.
- I think it's better to add copy_traits for both C and D for different data type. XE_2D_U16x8x16_ST_N here
means storing to D as F16, you shouldn't change it. If you want to store D as F32, you should add a new
GemmConfiguration for it. - can you refine the data type name in "FP16U4FP16F16FP16S4" according to your changes.
The data type " FP16U4FP16F16FP16S4 " in the name are: A, B, C, Mma, Scale, Zero
petercad
reviewed
Oct 24, 2025
| using CollectiveEpilogue = cutlass::epilogue::collective::CollectiveEpilogue< | ||
| EpilogueDispatchPolicy, | ||
| TileShape, | ||
| ElementAccumulator, |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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Outdated suggestions cannot be applied.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.
Fix for supporting different data type for
LinearCombination Output and Compute.