Initial support for tritongpu.upcast_mxfp lowering
#2951
Merged
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
Signed-off-by: Tiotto, Ettore <[email protected]>
Signed-off-by: Tiotto, Ettore <[email protected]>
etiotto
reviewed
Dec 6, 2024
third_party/intel/lib/TritonIntelGPUToLLVM/UpcastMXFPToLLVM.cpp
Outdated
Show resolved
Hide resolved
third_party/intel/lib/TritonIntelGPUTransforms/AccelerateMatmul.cpp
Outdated
Show resolved
Hide resolved
Signed-off-by: Tiotto, Ettore <[email protected]>
Signed-off-by: Tiotto, Ettore <[email protected]>
…m/intel/intel-xpu-backend-for-triton into liyang/upcast_mxfp_and_dot_scaled
Signed-off-by: Tiotto, Ettore <[email protected]>
Signed-off-by: Tiotto, Ettore <[email protected]>
etiotto
approved these changes
Dec 6, 2024
etiotto
changed the title
tritongpu.upcast_mxfp lowering and enable tl.dot_scaled unit teststritongpu.upcast_mxfp lowering
whitneywhtsang
approved these changes
Dec 6, 2024
| /// Given a 'DotOp' or 'ScaledDot' operation, return its DPAS engine type. | ||
| static DPASEngineType getDPASType(Operation *op); | ||
|
|
||
| // clang-format off |
Contributor
There was a problem hiding this comment.
why need to turn off clang-format?
Contributor
There was a problem hiding this comment.
because it formats the template horribly
Signed-off-by: Tiotto, Ettore <[email protected]>
Signed-off-by: Tiotto, Ettore <[email protected]>
anmyachev
pushed a commit
that referenced
this pull request
Dec 7, 2024
This PR adds initial codegen support for the new upcast_mxfp operation. Currently the codegen works when the source operand has blocked layout. This is a temporary limitation (we will want to support dot layout for that operand). Also, there is a failing test in test_core.py which, in this PR, is skipped. We will address that problem in a separate PR. --------- Co-authored-by: Tiotto, Ettore <[email protected]>
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 initial codegen support for the new upcast_mxfp operation. Currently the codegen works when the source operand has blocked layout. This is a temporary limitation (we will want to support dot layout for that operand).
Also, there is a failing test in test_core.py which, in this PR, is skipped. We will address that problem in a separate PR.