[Codegen, CUDA] Add FP8 Tensor Core Codegen #16950
Open
+324
−61
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Outdated suggestions cannot be applied.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.
Major changes of this pull request:
requires_cuda_compute_version
from 9 to 8.9 (since sm_89 ada architecture also supports fp8 tensor cores, which is the platform I have tested on).get_mma_intrin_group
andget_mma_intrin
functions, as the prior implementation assumed that input A and input B were of the same datatype. However, fp8 tensor cores can process combinations like e5m2e5m2, e5m2e4m3, e4m3e4m3, or e4m3e5m2. Note: This change may affect code in MLC that utilizesget_mma_intrin_group
.Check out the correctness:
expected output:
Please CC @yzh119