Skip to content

Pull requests: triton-lang/triton

New pull request New
215 Open 7,514 Closed
215 Open 7,514 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

[PROTON] Support perfetto_trace and clean up trace dump code
#10061 opened Apr 17, 2026 by Jokeren Contributor • Draft
[BACKEND] Extend support for small MMA: single 8x8x4 instructions.
#10060 opened Apr 16, 2026 by mwichro Loading…
5 of 7 tasks
1
[TritonGPU] Coalesce integer atomics
#10059 opened Apr 16, 2026 by jack-white1 Loading…
5 of 7 tasks
[AMD] f16_gemm gluon kernel improve pipeline
#10057 opened Apr 16, 2026 by guacamoleo Contributor • Draft
1
[WIP][AMD][TDM] Add partial TDM copy support via warp_bases
#10056 opened Apr 16, 2026 by jungpark-mlir Contributor • Draft
8
[FPSAN] Make scratch load/stores uncached, add cta barrier to global BarrierOp
#10055 opened Apr 16, 2026 by pawelszczerbuk Contributor Loading…
7
[GSan] Model TMAReduceOp as atomic
#10053 opened Apr 16, 2026 by peterbell10 Contributor Loading…
[Consan] Support CLC
#10052 opened Apr 16, 2026 by lezcano Contributor Loading…
2
[gluon][examples] MoE bmm1 in Gluon
#10047 opened Apr 15, 2026 by Mogball Collaborator Loading…
4
[FPSAN] Enable FPSan with combined instrumentation modes
#10045 opened Apr 15, 2026 by pawelszczerbuk Contributor Loading…
[AMD] Enable TDM with column major tensors and partitioned encodings
#10044 opened Apr 15, 2026 by plognjen Contributor • Draft
[Triton] Fix LLVMDILocalVariable pass crash with LLVM_EXTRACT_DI_LOCAL_VARIABLES
#10042 opened Apr 15, 2026 by byoshimi-gmail Contributor Loading…
5 of 7 tasks
2
[Backend] Bump to llvm/llvm-project@f0f96c7f
#10041 opened Apr 15, 2026 by antiagainst Member • Draft
Use TargetInfo for shared memory accesses
#10038 opened Apr 15, 2026 by neildhar Collaborator • Draft
Skip unnecessary boundary-check masks in RewriteTensorDescriptorToPointerPass
#10034 opened Apr 14, 2026 by gkmhub Loading…
1
1
[Nvidia][Gluon] Refactor convolution fprop kernel, add wgrad and dgrad kernels
#10030 opened Apr 14, 2026 by bingyizh233 Contributor Loading…
2
[AMD][DRAFT] PC Sampling, wave stall reasonings
#10020 opened Apr 13, 2026 by ZelboK Contributor • Draft
[DNR][PROTON][CUPTI] User-defined kernel activity records
#10018 opened Apr 13, 2026 by Jokeren Contributor • Draft
[Bugfix] Update truncation from i32 to i8 when storing to i1
#10011 opened Apr 13, 2026 by greatofdream Loading…
4 of 7 tasks
1
7
Add float8_e8m0fnu support to type canonicalization for dot_scaled
#10009 opened Apr 12, 2026 by GeisYaO Loading…
8
Fix nested def in @triton.jit to raise intended UnsupportedLanguageConstruct
#10004 opened Apr 11, 2026 by swjng Loading…
5 tasks done
Validate incompatible cache_modifier/eviction_policy combinations in NVIDIA backend
#10003 opened Apr 11, 2026 by swjng Loading…
5 tasks done
5
[NFC] Kill a couple helpers and make internal functions static
#9988 opened Apr 10, 2026 by lezcano Contributor Loading…
26
[DRAFT][DNR] Update FP16 warp-pipelined GEMM
#9978 opened Apr 9, 2026 by jungpark-mlir Contributor • Draft
[TritonGPU] Preserve memdesc_reshape encoding on type propagation
#9973 opened Apr 9, 2026 by Sibylau Loading…
5 of 7 tasks
15
ProTip! Add no:assignee to see everything that’s not assigned.