[ET-VK][runtime][2/3] Add typed dispatch size abstractions - #22289
Merged
Conversation
Pull Request resolved: #22224 Introduce `LocalWorkGroup`, `LwgShape`, `GlobalWorkGrid`, and `DispatchGridIntent` in `runtime/vk_api/DispatchGrid`. Remove legacy `WorkgroupSize`. Model local shapes as power-of-two exponents, support fitting against global extents, and validate local dimensions against device limits. Track dispatch intent independently from optional required local workgroup geometry. Authored with Codex. ghstack-source-id: 421281484 @exported-using-ghexport Differential Revision: [D117706351](https://our.internmc.facebook.com/intern/diff/D117706351/)
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/22289
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
Pull Request resolved: #22225 Migrate dispatch, prepack, profiling, and workgroup picker interfaces to `GlobalWorkGrid` and `LocalWorkGroup`. Wrap oversized linear grids using device limits and validate final geometry before `vkCmdDispatch`. Carry exact local geometry as `GlobalWorkGrid` metadata. Cooperative and reduction grid pickers set it explicitly, while `pick_required_lwg` validates and returns it. Preserve elongated Conv1d depthwise workgroups where the default shape would regress GPU latency. Add end-to-end oversized linear dispatch coverage. Authored with Codex. ghstack-source-id: 421281492 @exported-using-ghexport Differential Revision: [D117706355](https://our.internmc.facebook.com/intern/diff/D117706355/)
Add batch-aware direct, pointwise, depthwise, and im2col q8ta convolution dispatch. Batched im2col uses an NCHW scratch tensor and batch-strided shader indexing, with conservative groups=1 high-K/small-spatial routing bounded by a 32 MiB scratch cap. Grouped and large-spatial convolutions stay on the direct path. Authored with Codex. Differential Revision: [D117869783](https://our.internmc.facebook.com/intern/diff/D117869783/) ghstack-source-id: 421281491 Pull-Request: #22254
SS-JIA
approved these changes
Aug 29, 2026
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #22224 by @SS-JIA
^ Please use this as the source of truth for the PR details, comments, and reviews
ghstack PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/604/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/604/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/603/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/604/orig
Differential Revision: D117706351
@diff-train-skip-merge