Roofline: classify LATENCY_BOUND for small global-memory transfers - #962
Draft
ajassani wants to merge 1 commit into
Draft
Roofline: classify LATENCY_BOUND for small global-memory transfers#962ajassani wants to merge 1 commit into
ajassani wants to merge 1 commit into
Conversation
ajassani
requested review from
devalshahamd,
gabeweisz and
tsrikris
as code owners
August 24, 2026 20:44
ajassani
marked this pull request as draft
August 24, 2026 21:04
Codecov Report❌ Patch coverage is
📢 Thoughts on this report? Let us know! |
Extend global-memory roofline classification with a latency floor on small transfers. Legacy arch files without mem_latency_us keep the old bandwidth-only COMPUTE_BOUND/MEMORY_BOUND behavior and emit a one-time UserWarning when TreePerfAnalyzer is constructed. Co-authored-by: Cursor <cursoragent@cursor.com>
ajassani
force-pushed
the
ajassani/roofline-latency-bound
branch
from
August 24, 2026 21:30
00c9449 to
df25267
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
mem_latency_usin GPU arch JSON) so small transfers can be taggedLATENCY_BOUNDinstead of being lumped into bandwidth-boundMEMORY_BOUND.mem_latency_usis absent: bandwidth-onlyCOMPUTE_BOUND/MEMORY_BOUNDwith a one-timeUserWarningatTreePerfAnalyzerconstruction.examples/gpu_arch_example.mdand add unit/integration tests.Test plan
pytest tests/test_roofline_latency_bound.pypytest tests/test_roofline_bound.py