Skip to content

Pull requests: triton-lang/triton

New pull request New
160 Open 6,478 Closed
160 Open 6,478 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

Avoid using block_m=64 when having mxfp8 downcast in epilogue
#8830 opened Nov 25, 2025 by qnie-oai • Draft
7 tasks
[AMD] [ITT] Compilation crash on broadcasted registers
#8828 opened Nov 24, 2025 by alefimov-amd • Draft
refine lds layout order for transposed input
#8827 opened Nov 24, 2025 by scxiao • Draft
5 of 7 tasks
Fix AxisInfo rank mismatch for poison tensor pointers
#8824 opened Nov 24, 2025 by wdziurdz Loading…
2 tasks
1
2
[WIP][AMD] Extended FP conversion for gfx1250 arch
#8821 opened Nov 24, 2025 by ravil-mobile • Draft
1
[WIP] Enable ASan tests
#8819 opened Nov 24, 2025 by PMylon • Draft
add bailout for estimation and benchmarking in autotuning
#8816 opened Nov 24, 2025 by krillillin Loading…
5 of 7 tasks
4
[AMD] Turn TRITON_HIP_USE_ASYNC_COPY to be on by default
#8812 opened Nov 24, 2025 by antiagainst • Draft
[mxfp] allow non persistent mx matmul w/ strided layout
#8808 opened Nov 22, 2025 by jongsoo-openai • Draft
4 of 7 tasks
Respect disallow acc multibuffer in WS mode
#8802 opened Nov 21, 2025 by pawelszczerbuk Loading…
Make WarpSpecializePartitionsOp implement RegionBranchInterface
#8799 opened Nov 21, 2025 by neildhar Loading…
1
[WS] Support double buffering of scales in TMEM for tmem_store
#8795 opened Nov 21, 2025 by masahi Loading…
6
[WIP] Support JITFunction in preload
#8794 opened Nov 21, 2025 by ThomasRaoux • Draft
2
[AMD] Refactor kWidth Assignment
#8792 opened Nov 21, 2025 by christopherpriebe Loading…
4
Add update_tensor_descriptor operation to Triton/Gluon
#8786 opened Nov 20, 2025 by alexsamardzic Loading…
8
[Frontend] Support returning constexpr from jit functions
#8785 opened Nov 20, 2025 by peterbell10 Loading…
6
[AMD] Make arith.select handling in canonicalize pointers more Robust
#8779 opened Nov 20, 2025 by njriasan Loading…
2
[triton_kernels] Force matmul_ogs output to be TMA-compatible
#8774 opened Nov 19, 2025 by Maratyszcza Loading…
5 of 7 tasks
@ptillet
[mxfp] adjust num_stages for bf16/fp16 x mxfp
#8773 opened Nov 19, 2025 by jongsoo-openai Loading…
4 of 7 tasks
3
[Backend] Improve interval analysis in Membar
#8755 opened Nov 18, 2025 by nzaghen • Draft
16
Preserve constexprs returned from functions
#8749 opened Nov 17, 2025 by saagarjha Loading…
5 of 7 tasks
7
Add x scale swizzling for blackwell + batched matmul
#8747 opened Nov 17, 2025 by qnie-oai Loading…
5 of 7 tasks
6
[Doc] Add tables for the supported dtypes
#8746 opened Nov 17, 2025 by mattduerrmeier Loading…
5 of 7 tasks
CLC (cluster launch control) backend
#8740 opened Nov 16, 2025 by dshi7 Loading…
5
[SWP] Fix barrier location in loop lowering for MMA op with non-pipelined operands
#8732 opened Nov 14, 2025 by njriasan Loading…
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.