-
Notifications
You must be signed in to change notification settings - Fork 2.5k
Pull requests: triton-lang/triton
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[AMD][TRITON_KERNELS] enable expert parallelism implementation on AMD…
#9110
opened Dec 24, 2025 by
Liang-jianhao97
•
Draft
5 of 7 tasks
Improve error message for unsupported binary operators
#9100
opened Dec 24, 2025 by
Rudra-Tiwari-codes
Loading…
[LANGUAGE] change default 32-bit dot precision to TF32x3 (reland)
#9094
opened Dec 23, 2025 by
ptillet
Loading…
[BACKEND] Sink tmem allocs after pipelining to reduce liveranges
#9093
opened Dec 23, 2025 by
ThomasRaoux
Loading…
[BUILDER] Return operation from create_descriptor_store
#9091
opened Dec 23, 2025 by
mieshkiwrk
Loading…
Fix handling conflicting layouts when hoisting convert into conditionals
#9083
opened Dec 23, 2025 by
neildhar
Loading…
[NFC][gluon] Fix GluonSimplifyControlFlow's description
#9066
opened Dec 19, 2025 by
meshtag
Loading…
5 of 7 tasks
[Frontend] Fix setting attr on constexpr argument
#9053
opened Dec 19, 2025 by
lijinpei-amd
Loading…
[FRONTEND] Implement ragged TMA descriptor functionality in Gluon
#9045
opened Dec 18, 2025 by
alexsamardzic
Loading…
Support swiglu and gamma together in matmul
#9037
opened Dec 18, 2025 by
Bessss-zyw
Loading…
5 of 7 tasks
[BACKEND] Fix liveness bug in TritonGPURemoveLayoutConversions
#9036
opened Dec 18, 2025 by
njriasan
Loading…
Avoid out-of-bounds tl.store in
_ragged_tensor_metadata_memset kernel using mask
#9026
opened Dec 17, 2025 by
anmyachev
Loading…
Fix replaceUsesAndPropagateType helper
#9017
opened Dec 16, 2025 by
rupanshusoi
Loading…
5 of 7 tasks
[AMD] Use v_perm instruction for convert_layout acceleration
#9014
opened Dec 16, 2025 by
alefimov-amd
Loading…
Previous Next
ProTip!
Follow long discussions with comments:>50.