Skip to content

Pull requests: tile-ai/tilelang

New pull request New
52 Open 2,099 Closed
52 Open 2,099 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

[BugFix] Preserve buffer-loaded guard dependencies in pipelines
#3052 opened Aug 19, 2026 by kobecai Loading…
4 tasks done
1
4
[Misc] Remove incorrect ASF license headers from src files
#3051 opened Aug 19, 2026 by penguin-wwy Contributor Loading…
2
[BugFix][Layout] Preserve packed sub-byte byte ownership
#3046 opened Aug 17, 2026 by SamJSui • Draft
1
2
Add official compile-only CLI
#3045 opened Aug 17, 2026 by LibertychaserUS Loading…
1
8
[BugFix][CUDA] Support int4x2 and uint4x2 codegen
#3036 opened Aug 17, 2026 by SamJSui Loading…
2
5
[Language] Unify contiguous stride construction using a single implem…
#3016 opened Aug 16, 2026 by jjppp Contributor Loading…
1
2
[Transform] Preserve conditional type-cast store semantics
#2972 opened Aug 14, 2026 by GY-Bai Contributor • Draft
2
[Metal] Harden bf16 code generation and vectorization
#2971 opened Aug 14, 2026 by GY-Bai Contributor • Draft
2
[Metal] Make simdgroup synchronization tile-aware
#2970 opened Aug 14, 2026 by GY-Bai Contributor • Draft
2
[Metal] Harden eager adapter execution and argument binding
#2969 opened Aug 14, 2026 by GY-Bai Contributor • Draft
2
[Metal] Add reduction and infinity lowering
#2968 opened Aug 14, 2026 by GY-Bai Contributor • Draft
2
[BugFix][Hopper][Blackwell] Read the WGMMA/UMMA K-panel stride from the layout instead of the operand extent
#2965 opened Aug 14, 2026 by bigSheep123 Contributor Loading…
1
4
Add Windows ARM64 support to CI
#2962 opened Aug 13, 2026 by chinazhangchao Loading…
6
[BugFix][CUDA] Restore cp.async for predicated zero-fill copies
#2926 opened Aug 9, 2026 by hebo1221 Loading…
1
2
[Bugfix] Carry memory_order through vectorized atomic_add
#2924 opened Aug 9, 2026 by arcusbuilds Loading…
1
@SiriusNEO
16
[CI] Add gfx950 ROCm coverage on an MI350X runner
#2920 opened Aug 7, 2026 by andyluo7 Collaborator • Draft
2
[BugFix][CPU] Compute scalar GEMM products in accum dtype instead of input dtype
#2917 opened Aug 7, 2026 by Dino1844 Loading…
3
[ROCm] Scalarize vectorizable calls in vector Select lowering
#2915 opened Aug 7, 2026 by morluto Contributor Loading…
15
[BugFix] Reject layouts on shared buffers with non-constant extents
#2909 opened Aug 6, 2026 by sohampanda000 Loading…
1
12
[BugFix] Scalarize vectorized bool with more than 4 lanes
#2895 opened Aug 6, 2026 by jjppp Contributor Loading…
2
6
[BugFix][Transform] Preserve empty stepped loops during unrolling
#2887 opened Aug 5, 2026 by morluto Contributor Loading…
1
2
[BugFix] Reject arch-less CUDA target instead of silent sm_50 (#1049)
#2860 opened Aug 3, 2026 by cklxx Contributor • Draft
1
2
[Docs] Fix stale repository links
#2857 opened Aug 3, 2026 by morluto Contributor Loading…
6
[ROCm] Enable DeepSeek V4 FP8/FP4 activation quantization
#2847 opened Aug 3, 2026 by skyguan92 Contributor Loading…
@SiriusNEO
23
[BugFix][CUDA] Guard cp.async transfers by full source extent
#2842 opened Aug 1, 2026 by morluto Contributor Loading…
1
18
ProTip! Follow long discussions with comments:>50.