Skip to content

Pull requests: triton-lang/triton

New pull request New
62 Open 3,901 Closed
62 Open 3,901 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[BACKEND] Deprecate SharedToDotOperandMMAv2OrV3.cpp
#5734 opened Jan 28, 2025 by Jokeren Loading…
1
[NVIDIA] Prefer nvvm intrinsics over custom PTX
#5733 opened Jan 28, 2025 by peterbell10 Loading…
1
[AMD] AsyncCopyGlobalToLocal lowering to global.load.lds
#5729 opened Jan 28, 2025 by AlexAUT • Draft
[WIP][DNR][Pipeliner] Enable automatic loop fusion
#5726 opened Jan 28, 2025 by Mogball • Draft
1
Reapply "[Layouts] Propagate layouts into conditionals (#5610)"
#5725 opened Jan 28, 2025 by Mogball Loading…
8
[WIP] Support shared encoding defined with linear layout
#5720 opened Jan 27, 2025 by binarman • Draft
5
[AMD][BACKEND] Disable pingpong with non-local_load input.
#5718 opened Jan 27, 2025 by jungpark-mlir • Draft
7
Use torch.Tensor with unsigned types directly instead of TensorWrapper
#5715 opened Jan 27, 2025 by anmyachev Loading…
3
Bump actions/checkout from 3 to 4 dependencies Pull requests that update a dependency file
#5714 opened Jan 27, 2025 by dependabot bot Loading…
[Triton] Add top-k operation to Triton language
#5706 opened Jan 26, 2025 by peymanbr Loading…
5 of 7 tasks
6
Implement TMA store reduction for add, min, max
#5699 opened Jan 25, 2025 by ardaunal Loading…
5 of 7 tasks
8
[WIP][AMD] Add MFMA and WMMA layouts to LinearEncodingTest
#5698 opened Jan 24, 2025 by binarman • Draft
[AMD] Enable pingpong scheduling by default
#5696 opened Jan 24, 2025 by jungpark-mlir • Draft
7
Do not reorder transpose of dot operand that is used in ops other than dotOp
#5686 opened Jan 23, 2025 by pawelszczerbuk • Draft
1
[BACKEND] bump to llvm/llvm-project@c118864223c6
#5684 opened Jan 23, 2025 by makslevental Loading…
13
Fix assertion in ScanLowering for num_ctas>1
#5680 opened Jan 23, 2025 by safelix Loading…
5 of 7 tasks
4
[Proton][Dialect] Middle-end support of the Proton Dialect and the frontend Python package
#5677 opened Jan 23, 2025 by fywkevin • Draft
26
[WIP] [AMD] Remove "remove unsupported conversions" pass
#5674 opened Jan 22, 2025 by binarman • Draft
3
[LAYOUTS] Generalise HoistLayoutConversion to work with arbitrary layouts and chains of ops
#5673 opened Jan 22, 2025 by lezcano Loading…
6
[ANALYSIS][DEBUG] Output theoretical vs actual peak memory allocation size
#5658 opened Jan 21, 2025 by Jokeren Loading…
1
Ensured that dtype subclasses are hashable
#5657 opened Jan 20, 2025 by superbobry Loading…
4 of 7 tasks
[backend] NFC: Split architecture dependant and independant parts of FMA dot conversion
#5655 opened Jan 20, 2025 by binarman Loading…
5
Use dataframe index for x-axis in Mark class
#5614 opened Jan 15, 2025 by anmyachev • Draft
1
[Proton][Dialect] Add Proton Device Memory Buffer Init and Allocate Pass
#5606 opened Jan 14, 2025 by CRobeck Loading…
@fywkevin
19
[Frontend][Diagnostics] Improve emitting diagnostic information
#5581 opened Jan 11, 2025 by sfzhu93 Loading…
5 of 7 tasks
4
ProTip! Follow long discussions with comments:>50.