[webgpu] use u32 to represent f16 in uniform #25391
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
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR updates how 16-bit floats are represented in uniform buffers by packing them into 32-bit unsigned integers and adjusts buffer layout computation and WGSL codegen accordingly.
- Compute size and alignment for f16 uniforms as u32-backed containers in
webgpu_context.cc. - Generate WGSL bitcast expressions in
GetElementAtfor f16 access inshader_variable.h. - Change uniform struct declarations to use u32 types and adjusted lengths in
shader_helper.cc.
Reviewed Changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.
| File | Description |
|---|---|
| onnxruntime/core/providers/webgpu/webgpu_context.cc | Revised uniform offset/size logic to handle f16 as u32 with correct alignment and padding. |
| onnxruntime/core/providers/webgpu/shader_variable.h | Expanded GetElementAt to emit bitcast<vec2<f16>> accesses for f16 uniforms. |
| onnxruntime/core/providers/webgpu/shader_helper.cc | Mutates data_type/length for f16 and emits appropriate WGSL types (u32, vecN, array). |
Comments suppressed due to low confidence (1)
onnxruntime/core/providers/webgpu/webgpu_context.cc:381
- The comment for the f16 array threshold (>8) does not match the implementation (which uses
length > 6). Please update the comment to reflect the actual branch condition or adjust the threshold to align with the intended behavior.
// - length > 8 : array<vec4<u32>, N> (align 16) (size 16 * N, N = ceil(length / 8))
vraspar
approved these changes
Jul 30, 2025
Contributor
vraspar
left a comment
vraspar
left a comment
There was a problem hiding this comment.
I wonder how will it impact perf where device support f16, but we use u32
Contributor
Author
technically the perf impact is Anyway, we haven't got any actual use case for f16 in uniform yet, so this is not going to impact anything. |
adrianlizarraga
pushed a commit
that referenced
this pull request
Aug 8, 2025
### Description For f16 uniform variables, use u32 to bit-wise represent them. ### Motivation and Context Some devices supports f16 in shader/storage buffer, but not in uniform buffers. Dawn will set the f16_support to false for them. However, we don't necessarily have to use f16 in uniform. This change together with #25349 will enable using f16 models on some Android devices.
adrianlizarraga
added a commit
that referenced
this pull request
Aug 8, 2025
…5, 25652 (#25701) ### Description Cherry-pick the following PRs into the `rel-1.23.0` branch: - #25391 - #25611 - #25656 - #25346 - #25374 - #25664 - #25675 - #25652 ### Motivation and Context <!-- - Why is this change required? What problem does it solve? - If it fixes an open issue, please link to the issue here. --> --------- Co-authored-by: Yulong Wang <[email protected]> Co-authored-by: Ishwar Raut <[email protected]> Co-authored-by: Maximilian Müller <[email protected]> Co-authored-by: Gaurav Garg <[email protected]> Co-authored-by: Scott McKay <[email protected]> Co-authored-by: Chi Lo <[email protected]> Co-authored-by: Abhishek Jindal <[email protected]> Co-authored-by: Dmitri Smirnov <[email protected]>
sanketkaleoss
pushed a commit
to sanketkaleoss/onnxruntime
that referenced
this pull request
Aug 11, 2025
### Description For f16 uniform variables, use u32 to bit-wise represent them. ### Motivation and Context Some devices supports f16 in shader/storage buffer, but not in uniform buffers. Dawn will set the f16_support to false for them. However, we don't necessarily have to use f16 in uniform. This change together with microsoft#25349 will enable using f16 models on some Android devices.
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.
Description
For f16 uniform variables, use u32 to bit-wise represent them.
Motivation and Context
Some devices supports f16 in shader/storage buffer, but not in uniform buffers. Dawn will set the f16_support to false for them. However, we don't necessarily have to use f16 in uniform.
This change together with #25349 will enable using f16 models on some Android devices.