Repository navigation
Add the Base/SharedMemory tutorial - #184
Merged
Merged
Conversation
WeiqunZhang
force-pushed
the
shared-memory
branch
from
November 4, 2026 19:04
f38cf09 to
2a8345f
Compare
Bin the cells by one field and sum a signed integer field over each bin, with ParallelFor and global Long atomics, and with a hand-written kernel launched by amrex::launch that keeps private int sums per thread block in shared memory (CUDA/HIP and SYCL bodies), with a separate CPU path using private sums per OpenMP thread. Atomic adds to the same global address are serialized, so the global version slows down as the bins get fewer. The results are compared and timed with BL_PROFILE.
WeiqunZhang
force-pushed
the
shared-memory
branch
from
November 4, 2026 19:59
2a8345f to
d169933
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Bin the cells by one field and sum a signed integer field over each bin, with ParallelFor and global Long atomics, and with a hand-written kernel launched by amrex::launch that keeps private int sums per thread block in shared memory (CUDA/HIP and SYCL bodies), with a separate CPU path using private sums per OpenMP thread. Long atomics are compare-and-swap loops on CUDA and HIP, which makes the contended global version slow. The results are compared and timed with BL_PROFILE.