Single-pass scan algorithm - #157
Draft
AntonReinhard wants to merge 9 commits into
Draft
Conversation
psychocoderHPC
requested changes
Jun 27, 2025
AntonReinhard
force-pushed
the
scan_single_pass
branch
from
August 11, 2025 13:31
febe826 to
3a33b8e
Compare
AntonReinhard
force-pushed
the
scan_single_pass
branch
4 times, most recently
from
August 13, 2025 16:19
e581697 to
f0cd0bc
Compare
Collaborator
Author
|
Rebased on #214 to see if it works |
AntonReinhard
force-pushed
the
scan_single_pass
branch
from
August 14, 2025 15:36
f0cd0bc to
fb32717
Compare
AntonReinhard
force-pushed
the
scan_single_pass
branch
from
September 2, 2025 14:07
fb32717 to
bc78cbf
Compare
AntonReinhard
force-pushed
the
scan_single_pass
branch
2 times, most recently
from
December 18, 2025 10:32
dd51878 to
bed39af
Compare
AntonReinhard
force-pushed
the
scan_single_pass
branch
from
December 18, 2025 18:09
bed39af to
665fde8
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Work in Progress
The algorithm is correct (at least for CPU and CUDA), but currently uses non-Alpaka memory fences (because they're not yet implemented) and is also slow, because the exclusive block prefix calculation is slow.
If the exclusive prefix calculation is left out, the performance on CUDA is roughly the same as the cuda library implementation, which is promising, since it should be possible to almost entirely hide the latency of this calculation.
Based on https://research.nvidia.com/sites/default/files/pubs/2016-03_Single-pass-Parallel-Prefix/nvr-2016-002.pdf