[ET-VK][qconv] Add flexible layout impl for im2col #17268
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/17268
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
SS-JIA
force-pushed
the
gh/SS-JIA/410/orig
branch
from
6876097 to
3eb0bd4
Compare
Pull Request resolved: #17249 This implements an im2col-based approach for quantized conv2d, which transforms convolution into matrix multiplication. The im2col transformation extracts sliding windows from the input tensor and reshapes them into a 2D matrix, enabling reuse of the optimized pointwise convolution shader for the compute-intensive portion. Two im2col shaders are added: - `q8ta_im2col.glsl`: Generic shader with layout-agnostic input access via BufferMetadata and specialization constants - `q8ta_im2col_4w4c.glsl`: Optimized shader for 4W4C input layout that exploits the alignment between consecutive width positions and packed channel values The im2col output is always stored in 4W4C layout to match the expected input format of the pointwise convolution shader. The operator is registered as `etvk.q8ta_conv2d_im2col.default` and currently supports non-grouped convolutions where input channels is a multiple of 4. Authored with assistance from Claude. ghstack-source-id: 338638552 @exported-using-ghexport Differential Revision: [D92407723](https://our.internmc.facebook.com/intern/diff/D92407723/)
SS-JIA
force-pushed
the
gh/SS-JIA/411/orig
branch
from
d49bf74 to
5e545ea
Compare
SS-JIA
approved these changes
Feb 6, 2026
This PR needs a
|
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: #17249 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/411/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/411/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/SS-JIA/410/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/SS-JIA/411/orig
Differential Revision: D92407723
@diff-train-skip-merge