webgpu: enlarge the splitted dimInner size #6755
Merged
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
xhcao
approved these changes
Aug 17, 2022
| let numTiles = ${splitK ? '1' : '(uniforms.dimInner - 1) / TileInner + 1'}; | ||
| var kStart = ${splitK ? 'i32(globalId.z) * TileInner' : '0'}; | ||
| let numTiles = ${ | ||
| splitK ? `${splitedDimInner / tileInner}` : |
There was a problem hiding this comment.
Nit. Do we need to change it to (splitedDimInner - 1) / TileInner + 1 if splitedDimInner % tileInner is not equal to 0
| @@ -323,8 +326,9 @@ export function makeMatMulPackedSource( | |||
| let globalRowStart = i32(workgroupId.y) * ${tileAOuter}; | |||
|
|
|||
| let numTiles = ${ | |||
| splitK ? '1' : '(uniforms.dimInner - 1) / TileInner + 1'}; | |||
| var kStart = ${splitK ? 'i32(globalId.z) * TileInner' : '0'}; | |||
| splitK ? `${splitedDimInner / tileInner}` : | |||
There was a problem hiding this comment.
Change it to Math.ceil(splitedDimInner / tileInner). Thanks.
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Outdated suggestions cannot be applied.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.
This PR enlarges the splitted dimInner size to reduce the number of calls to atomic operations.
Before
Conv2d 0.28ms input0: [5,5,2048] input1:[1,1,2048,32]
After
conv2d 0.12ms input0: [5,5,2048] input1:[1,1,2048,32]
To see the logs from the Cloud Build CI, please join either our discussion or announcement mailing list.
This change is