Skip to content

refactor(examples): use SMEM layout selectors - #17

Merged
ArthurinRUC merged 1 commit into
mainfrom
refactor/smem-layout-selectors
May 19, 2026
Merged

ArthurinRUC merged 1 commit into
mainfrom
refactor/smem-layout-selectors

Conversation

@ArthurinRUC

Copy link
Copy Markdown
Owner

Summary

  • Replace hardcoded Hopper K_SW128 SMEM atoms in examples 11-14 with selector-based layout selection.
  • Use CUTLASS ss_smem_selector in C++ and hopper_helpers.get_smem_layout_atom in CuTe DSL.
  • Expand example 11 comments with byte_offset vs element_offset and Swizzle<3,4,3> vs 16-bit Swizzle<3,3,3> examples.

Validation

  • Local: ruff check, py_compile, clang-format -n -Werror, git diff --check.
  • G2-docker smoke: examples 11-14 C++ and CuTe DSL paths, including warpgroup_mma_multicast.py and warp_specialization_pipeline.py, all PASS at M=256, N=256, K=64.

Replace hardcoded Hopper K_SW128 shared-memory atoms in examples 11-14 with CUTLASS/CuTe selector logic. Expand the example 11 comment to explain byte-address swizzling, the 128B Swizzle<3,4,3> canonical form, and the 16-bit Swizzle<3,3,3> element-offset equivalence.
@ArthurinRUC
ArthurinRUC merged commit 43ecce7 into main May 19, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant