Update dpnp.fft to run on CUDA
#2332
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
vlad-perevezentsev
requested review from
antonwolfy,
AlexanderKalistratov,
vtavana and
ndgrigorian
as code owners
|
View rendered docs @ https://intelpython.github.io/dpnp/index.html |
|
Array API standard conformance tests for dpnp=0.17.0dev7=py312he4f9c94_19 ran successfully. |
antonwolfy
reviewed
Feb 24, 2025
antonwolfy
approved these changes
Feb 24, 2025
github-actions bot
added a commit
that referenced
this pull request
Feb 24, 2025
This PR suggests fixing the current issues with `dpnp.fft.fftn()` and `dpnp.fft.rfftn()` on CUDA and removing the skip tests for them The `incorrect result` issue for `dpnp.fft.fftn()` on cuda was because preparing the input array when `batch_fft=True` could change it to `F contiguous` array. cuFFT for correct execution requires `C contiguous` array as input. The issue with raising `Invalid strides` error is a bug in oneMath ([631](uxlfoundation/oneMath#631)). As a workaround until this issue is solved it is suggested to use swap of the last two axes if the last dimension is 1 and there are multiple axes. In this case the strides inside cuFFT are calculated correctly. Additionally updated the arguments for `test_erf` in `skipped_tests_cuda.tbl` to skip it on cuda. 8efd438
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Outdated suggestions cannot be applied.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 suggests fixing the current issues with
dpnp.fft.fftn()anddpnp.fft.rfftn()on CUDA and removing the skip tests for themThe
incorrect resultissue fordpnp.fft.fftn()on cuda was because preparing the input array whenbatch_fft=Truecould change it toF contiguousarray. cuFFT for correct execution requiresC contiguousarray as input.The issue with raising
Invalid strideserror is a bug in oneMath (631).As a workaround until this issue is solved it is suggested to use swap of the last two axes if the last dimension is 1 and there are multiple axes.
In this case the strides inside cuFFT are calculated correctly.
Additionally updated the arguments for
test_erfinskipped_tests_cuda.tblto skip it on cuda.