feat(cudf): Add velox-cudf support for TopN #14797
Open
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
karthikeyann
requested review from
bdice,
devavret and
mhaseeb123
as code owners
meta-cla
bot
added
the
CLA Signed
✅ Deploy Preview for meta-velox canceled.
|
karthikeyann
changed the title
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
karthikeyann
force-pushed
the
fea-CudfTopN
branch
from
0bfa61b to
2288ca9
Compare
Co-authored-by: Copilot <[email protected]>
There was a problem hiding this comment.
Pull Request Overview
This PR adds GPU-accelerated TopN operator support to velox-cudf integration. The implementation provides a CudfTopN operator that executes TopN operations on GPU using cuDF libraries, with comprehensive test coverage to verify correctness against CPU implementations.
- Implements CudfTopN operator with batch-based optimization for managing memory and performance
- Adds comprehensive test suite covering various TopN scenarios including multi-key sorting, filtering, and edge cases
- Includes a bug fix for CudfFilterProject to properly handle null values in filter conditions
Reviewed Changes
Copilot reviewed 7 out of 7 changed files in this pull request and generated 3 comments.
Show a summary per file
| File | Description |
|---|---|
| velox/experimental/cudf/exec/CudfTopN.h | Header file defining the CudfTopN operator class with batch management strategy |
| velox/experimental/cudf/exec/CudfTopN.cpp | Implementation of GPU-accelerated TopN operations using cuDF sorting APIs |
| velox/experimental/cudf/exec/ToCudf.cpp | Integration of CudfTopN into the operator replacement framework |
| velox/experimental/cudf/tests/TopNTest.cpp | Comprehensive test suite for TopN functionality verification |
| velox/experimental/cudf/tests/CMakeLists.txt | Build configuration for TopN tests |
| velox/experimental/cudf/exec/CMakeLists.txt | Build configuration including CudfTopN source |
| velox/experimental/cudf/exec/CudfFilterProject.cpp | Bug fix for null handling in filter conditions |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Co-authored-by: Copilot <[email protected]>
karthikeyann
force-pushed
the
fea-CudfTopN
branch
from
39e6aad to
1e100cd
Compare
Co-authored-by: Copilot <[email protected]>
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 adds operator CudfTopN to run TopN in GPU.
Unit tests are added to verify the output.
The CudfTopN stores top N values of each input batch, and if number of batches exceed 5, the batches are concatenated together and topN values are stored back to the batch vector. getOutput will concatenated all batches together and return topN values.
The tests are same as TopN, but Setup makes sure they run in GPU.
This PR has additional fix 2288ca9 for cudfFilterProject which affects filter results with null values.