[ET-VK] Minor performance improvements to native layer norm. #9892
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
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) [ghstack-poisoned]
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/9892
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit cfb8351 with merge base 1facfa9 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
trivedivivek
added a commit
that referenced
this pull request
Apr 4, 2025
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) ghstack-source-id: 276053981 Pull Request resolved: #9892
facebook-github-bot
added
the
CLA Signed
|
This pull request was exported from Phabricator. Differential Revision: D72430290 |
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) [ghstack-poisoned]
trivedivivek
added a commit
that referenced
this pull request
Apr 7, 2025
Pull Request resolved: #9892 This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) ghstack-source-id: 276439596
|
This pull request was exported from Phabricator. Differential Revision: D72430290 |
SS-JIA
approved these changes
Apr 7, 2025
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) [ghstack-poisoned]
trivedivivek
added a commit
that referenced
this pull request
Apr 7, 2025
Pull Request resolved: #9892 This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. ghstack-source-id: 276575089 Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/)
|
This pull request was exported from Phabricator. Differential Revision: D72430290 |
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/) [ghstack-poisoned]
trivedivivek
added a commit
that referenced
this pull request
Apr 8, 2025
Pull Request resolved: #9892 This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch. In this new approach: The mean and variance values are calculated in 2 separate passes. Shader is dispatched based on input texture size, and input texel is read and stored in shared memory. Input stored in shard memory is then summed up using a reduce function. This implementation better utilizes a GPUs parallel processing capabilities. ghstack-source-id: 276877983 Differential Revision: [D72430290](https://our.internmc.facebook.com/intern/diff/D72430290/)
|
This pull request was exported from Phabricator. Differential Revision: D72430290 |
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.
Stack from ghstack (oldest at bottom):
This diff introduces minor performance improvements to the native layer norm function in the Vulkan backend of Executorch.
In this new approach:
The mean and variance values are calculated in 2 separate passes.
Shader is dispatched based on input texture size, and input texel is read and stored in shared memory.
Input stored in shard memory is then summed up using a reduce function.
This implementation better utilizes a GPUs parallel processing capabilities.
Differential Revision: D72430290