Make VNNI more robust #1001
Merged
Make VNNI more robust #1001
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
|
No performance regression. |
rengolin
reviewed
Jan 16, 2025
rolfmorel
reviewed
Jan 16, 2025
adam-smnk
added
benchmark
rengolin
approved these changes
Jan 17, 2025
|
The API has been improved as suggested. Well, we'll still see in practice if it's good but already works better after this cleanup. I will follow up with further cleanup to reenable integration tests. That might require more relaxed verification and probably improvements to some test cases. But let's do it separately to weigh tradeoffs properly. Otherwise, should be fine. |
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Outdated suggestions cannot be applied.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.
Improves VNNI validation and lit tests to be target independent w.r.t. VNNI blocking factor.
VNNI infrastructure is extended with DLTI support that allows to override target's default VNNI factor.
This allows to fix VNNI shape for lit testing to ensure that general pass logic works fine. This is primarily applied to tests that use prepacked VNNI shapes.
When possible, tests' checks are generalized to account for varying packing factor.
Test shapes are adjusted to prevent failure due to incompatible VNNI packing sizes compared to dimension size used for verification.
BF16 integration tests are now disabled on non-x86 systems to avoid runtime mismatch between prepacked shapes and microkernel requirements.