Conversation
Contributor
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
This was referenced Sep 21, 2026
tpn
force-pushed
the
codex/cuda-coop-cutlass-06-exchange-shuffle
branch
from
September 21, 2026 04:15
5bb212c to
94a5bcf
Compare
tpn
force-pushed
the
codex/cuda-coop-cutlass-07-guides
branch
from
September 21, 2026 04:15
667646e to
9a387eb
Compare
Give Numba-CUDA-MLIR and CUTLASS explicit programming and developer entry points. Keep shared concepts and backend coverage in the overview, and call out qualified API differences in each backend guide. Add a CUTLASS compiler walkthrough and preserve existing page paths, section bookmarks, and executable examples. Signed-off-by: Trent Nelson <trent@trent.me>
tpn
force-pushed
the
codex/cuda-coop-cutlass-06-exchange-shuffle
branch
from
September 21, 2026 04:31
94a5bcf to
dd75f17
Compare
tpn
force-pushed
the
codex/cuda-coop-cutlass-07-guides
branch
from
September 21, 2026 04:31
9a387eb to
950616c
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Prerequisite: #11532.
Readers now choose an explicit Programming Guide or Developer Guide for
Numba-CUDA-MLIR or CUTLASS from a shared
cuda.coopoverview. Shared conceptsand namespace contracts remain in the overview; each programming guide
explains how to write kernels and identifies qualified API differences.
The new CUTLASS Developer Guide follows an executable tile copy through
compiler selection, CUB planning, provider generation, LTO linking, and
scratch allocation. It explains the required compiler capabilities, exact
launch facts, failure recovery, and checks for changes to the integration.
Existing page paths and section bookmarks remain available.
This foundation documents CUTLASS coverage through Exchange and Shuffle.
Merge Sort, Radix Sort/Rank, and TopK are marked as pending here and become
available in the following implementation layers. The later documentation expansion PR
adds the debugger walkthrough and expanded reference material.
Validation
Focused Sphinx builds and checks of links, bookmarks, and executable excerpts
passed during development. The prior 12-layer tip also passed the full Sphinx
build with only three existing libcu++ warnings, plus browser checks of all
four guides and their navigation. These are cumulative checks of the prior 12-layer
tip; this intermediate documentation head has not been separately rebuilt
on the refreshed base.
Final integrated regression: Passed: 879 shared-contract checks, 1,415 Numba host/compiler checks, 798 Numba GPU checks, seven standalone Numba commands, 1,757 CUTLASS host/compiler checks, and 2,136 CUTLASS GPU/compiler-lifecycle checks. Both GPU suites ran with race checking enabled; there were no failures or skips.
Final installed-wheel qualification: Fresh-wheel checks passed: all 252 Python sources/stubs match the final source, 40 packaging/typing tests passed, and base-only imports plus four mixed-backend activation cases passed. Installed GPU consumers passed 20 common/qualified example runs, four storage variants, runtime-first and disabled-auto-registration modes, and three executable guide tests.
Final documentation qualification: Full Sphinx rendered with only three existing libcu++ warnings; focused cooperative pages build without warnings. API signatures, 39 source excerpts, downloads, anchors, and local links pass. Local and live browser checks passed for 32 CUTLASS objects and all 14 explorers. The published docs match the tested source; 396 served resources match build hashes and the separate Numba docs remain unchanged.
Stack
Part 8 of 16 above #11392. Previous: #11532 · Next: #11534.
Complete stack and current common API coverage: #11541.