Skip to content

Pull requests: triton-lang/triton

New pull request New
71 Open 4,214 Closed
71 Open 4,214 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

Cap warp count to 16 for devices with warp size 64
#6125 opened Mar 6, 2025 by schung-amd Loading…
5 of 7 tasks
2
[WIP][NFC] Hoist non-dot-encoding-sensitive inferOpEncoding to base class
#6121 opened Mar 6, 2025 by LiyangLingIntel • Draft
3
Support NVIDIA 50 series GPU
#6120 opened Mar 6, 2025 by JiamingMai Loading…
5 of 7 tasks
3
[WIP][Remarks][Coalesce] Emit Remarks on divisibility when the number of element per thread converges to one
#6118 opened Mar 6, 2025 by sfzhu93 • Draft
7 tasks
[Refactor] Remove getVectorSize and getContiguity functions due to incorrect behavior
#6116 opened Mar 5, 2025 by joviliast • Draft
[PROTON] Added custom metrics expressions
#6115 opened Mar 5, 2025 by jleeleee Loading…
2 tasks
[Release 3.3] [CP] [AMD] Disable f16/bf16 buffer atomic operations (#6090)
#6112 opened Mar 5, 2025 by jataylo Loading…
[Release 3.3] [CP] Improve JIT runtime by 30%
#6111 opened Mar 5, 2025 by jataylo Loading…
[AMD] Add Fp32<->OCP Fp8/Bf8 conversions on mi350
#6110 opened Mar 5, 2025 by yiqian1 • Draft
9
[AMD] Restrict Ping Pong Scheduler Movable Memory Ops to Dot Product Contributors
#6108 opened Mar 5, 2025 by njriasan Loading…
5 of 7 tasks
1
Fix F16 -> F8E5M2 RTNE conversion logic on A100.
#6102 opened Mar 4, 2025 by bchetioui Loading…
5 tasks done
2
[AMD] Rework load-store redundant data masking
#6101 opened Mar 4, 2025 by plognjen • Draft
1
[PROTON-DEV] Implement proton gpu global scratch alloc to llvm lowering
#6091 opened Mar 3, 2025 by CRobeck Loading…
[WIP][AMD]Enable Tests for Transposed B for FP8/BF8 dot_scaled
#6078 opened Mar 1, 2025 by knwng • Draft
3
[PIPELINE] AssignLatencies for mmav5
#6077 opened Mar 1, 2025 by pawelszczerbuk Loading…
2
[WIP][AMD][OPTIMIZER] Optimize transposed GEMM operand in shared memory
#6074 opened Feb 28, 2025 by binarman • Draft
[Optimizer] Skip Linear encoding in canonicalization
#6071 opened Feb 28, 2025 by binarman • Draft
2
[LAYOUTS] Get warp number and thread number from Module
#6068 opened Feb 28, 2025 by lezcano Loading…
[do_bench] Reduce host overhead impact in do_bench
#6066 opened Feb 28, 2025 by AlexAUT Loading…
2
Emit remarks for ReduceOp lowering failing to fit within a single warp
#6062 opened Feb 28, 2025 by plotfi Loading…
1
Support creating int1 tensors in interpreter
#6058 opened Feb 28, 2025 by saagarjha Loading…
5 of 7 tasks
4
[WIP][DNR] Loop partitioning
#6053 opened Feb 27, 2025 by Mogball • Draft
Add HoistLoad pass
#6051 opened Feb 27, 2025 by ardaunal Loading…
4 of 7 tasks
38
[AMD][Pipeline] Introduce compute-depth based pipelining
#6048 opened Feb 27, 2025 by sjw36 • Draft
[NVIDIA] Propagate TMA encoding from MMA operand
#6045 opened Feb 27, 2025 by peterbell10 • Draft
9
ProTip! Follow long discussions with comments:>50.