NFFT on GPU based on Package Extensions and GPUArrays.jl #136
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
nHackel
commented
Jun 27, 2024
| U = params.storeDeconvolutionIdx ? N : ntuple(d->0,D) | ||
| tmpVecHat = adapt(arr, zeros(Complex{T}, U)) | ||
|
|
||
| deconvIdx = adapt(arr, Int32.(deconvolveIdx)) |
There was a problem hiding this comment.
Here I changed the indices to be Int32, as these are usually more efficient for a GPU
nHackel
commented
Jun 27, 2024
|
|
||
| deconvIdx = adapt(arr, Int32.(deconvolveIdx)) | ||
| winHatInvLUT = adapt(arr, windowHatInvLUT[1]) | ||
| B_ = adapt(arr, Complex{T}.(B)) # Bit hacky |
There was a problem hiding this comment.
In my test B had an eltype of Float32 and I didn't see a way where I could tell adapt to change the eltype. So I changed it beforehand
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #136 +/- ##
==========================================
- Coverage 88.63% 88.28% -0.35%
==========================================
Files 10 12 +2
Lines 1126 1195 +69
==========================================
+ Hits 998 1055 +57
- Misses 128 140 +12 ☔ View full report in Codecov by Sentry. |
Merged
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Outdated suggestions cannot be applied.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 a new NFFT plan which is essentially just a copy of the current CuNFFT Plan. However, in this PR the plan is writing in a GPU-agnostic way and will work for any GPU implementing an FFT-Plan.
The plan can also be tested with JLArrays in CI without requiring a GPU