Skip to content

Pull requests: NVIDIA/Fuser

New pull request New
211 Open 4,695 Closed
211 Open 4,695 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

add tma transpose auto scheduler
#5982 opened Feb 19, 2026 by liqiangxl • Draft
Use std::optional instead of direct default argument for LaunchParams Direct Bindings Python extension with direct mapping to NvFuser CPP objects.
#5980 opened Feb 18, 2026 by rdspring1 Loading…
8
split transpose scheduler into tma (place-holder) and non-tma versions
#5978 opened Feb 18, 2026 by liqiangxl Loading…
6
Refactor getCommunicationInfo: extract getCommunicationInfoForParallelType
#5977 opened Feb 18, 2026 by wujingyue Loading…
18
add manual scheduling of transpose fusion with tma
#5976 opened Feb 18, 2026 by liqiangxl Loading…
5
[IR Container] Phase 2 IR Container Refactor
#5975 opened Feb 18, 2026 by mdavis36 • Draft
use relative indices in post methods
#5974 opened Feb 18, 2026 by Priya2698 Loading…
30
Don't decompose SqueezeOp that can be lowered to communication
#5972 opened Feb 18, 2026 by wujingyue • Draft
3
[IR Container] Phase 2.6 Concurrency & Thread Safety
#5971 opened Feb 18, 2026 by mdavis36 Loading…
8
Fix getCommunicationInfo for multi-dimensional meshes
#5969 opened Feb 18, 2026 by wujingyue Loading…
1
7
skip bcast domains in transpose scheduler mapped domain check
#5967 opened Feb 17, 2026 by liqiangxl Loading…
5
Add TMA scheduler for outer2D reduction
#5966 opened Feb 17, 2026 by tbqh Loading…
27
[IR Container] Phase 2.5 Copy-Move Semantics
#5964 opened Feb 12, 2026 by mdavis36 Loading…
10
CollectivePermute based allgather in host for-loop
#5963 opened Feb 12, 2026 by Priya2698 • Draft
2
[IR Container] Phase 2.4 Per-fusion statement tracking
#5961 opened Feb 12, 2026 by mdavis36 Loading…
9
[IR Container] Phase 2.3 Basic shared ptr
#5960 opened Feb 12, 2026 by mdavis36 Loading…
6
[IR Container] Phase 2.2 Fusion Statement Registration
#5958 opened Feb 12, 2026 by mdavis36 Loading…
5
[IR Container] Phase 2.1 Ownership Fusion Special Vals & Axioms
#5954 opened Feb 11, 2026 by mdavis36 Loading…
8
Llu/transpose tma
#5945 opened Feb 9, 2026 by liqiangxl • Draft
Use the default/cached value when a BuildConfig field is None
#5943 opened Feb 9, 2026 by wujingyue Loading…
1
6
remove input cached smem bank conflicts in transpose scheduler
#5930 opened Feb 6, 2026 by liqiangxl • Draft
9
increase transpose tile size to meet required bytes in flight
#5928 opened Feb 6, 2026 by liqiangxl • Draft
2
Broadcast-based allgather in host for-loop
#5925 opened Feb 6, 2026 by Priya2698 Loading…
45
Reduce-based MM+RS in MultiDeviceExecutor
#5923 opened Feb 5, 2026 by nsarka Loading…
@nsarka
35
Remove legacy bindings Direct Bindings Python extension with direct mapping to NvFuser CPP objects.
#5917 opened Feb 3, 2026 by rdspring1 Loading…
26
ProTip! Add no:assignee to see everything that’s not assigned.