[GPU] Take element type bitwidth into account with vector size. #19987
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
nirvedhmeshram
requested review from
MaheshRavishankar,
qedawkins,
kuhar,
Groverkss and
antiagainst
as code owners
nirvedhmeshram
force-pushed
the
elementwise_tile_size2
branch
3 times, most recently
from
97c9fcf to
b38a7bd
Compare
|
Could you add a test for this https://gist.github.com/MaheshRavishankar/2127ecdbc742ef11b98b55e522c6dce8 |
compiler/src/iree/compiler/Codegen/Dialect/GPU/TargetUtils/ConfigUtils.cpp
Outdated
Show resolved
Hide resolved
nirvedhmeshram
force-pushed
the
elementwise_tile_size2
branch
from
56e6f82 to
10ce5c6
Compare
Co-authored-by: MaheshRavishankar <[email protected]> Signed-off-by: MaheshRavishankar <[email protected]> Signed-off-by: Nirvedh <[email protected]>
Signed-off-by: Nirvedh <[email protected]>
nirvedhmeshram
force-pushed
the
elementwise_tile_size2
branch
from
10ce5c6 to
1f23d9f
Compare
Signed-off-by: Nirvedh <[email protected]>
nirvedhmeshram
force-pushed
the
elementwise_tile_size2
branch
from
1f23d9f to
6485474
Compare
Signed-off-by: Nirvedh <[email protected]>
nirvedhmeshram
force-pushed
the
elementwise_tile_size2
branch
from
331092b to
c26d3da
Compare
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Outdated suggestions cannot be applied.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 is for TileAndFuse workgroup and thread distribution logic. Before this PR we only tried to vectorize to size 4 but now we vectorize to a bitWidth of 128. This can however result in no vectorization for small dimensions so the PR also introduces a fallback for smaller shapes to vector size of 4.
Co-authored-by: MaheshRavishankar [email protected]
Signed-off-by: MaheshRavishankar [email protected]
Signed-off-by: Nirvedh [email protected]