Skip to content

Pull requests: NVIDIA/Fuser

New pull request New
162 Open 3,614 Closed
162 Open 3,614 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

Vectorize fp4 even if the last dim is not contiguous
#4669 opened Jun 24, 2025 by zasdfgbnm Loading…
2
FP4 support for discontiguous tensors
#4668 opened Jun 24, 2025 by zasdfgbnm Loading…
2
FP4 support in pointwise scheduler
#4666 opened Jun 24, 2025 by zasdfgbnm • Draft
1
Add new data type Float8_e8m0fnu
#4665 opened Jun 24, 2025 by zasdfgbnm Loading…
3
fix warp specialized tma for ln bwd
#4663 opened Jun 24, 2025 by liqiangxl Loading…
3
Create nvfp4 grouped gemm bindings in direct bindings Cutlass Matmuls Thunder-Inference-Demo
#4662 opened Jun 24, 2025 by rdspring1 • Draft
2 tasks
2
Removing the 2D assumption of index put accumulate
#4660 opened Jun 19, 2025 by naoyam Loading…
6
Unify propagation rules in propagateShardingsPass
#4652 opened Jun 17, 2025 by Priya2698 Loading…
14
Host IR LLVM Lowering 1: Build Config Change & Initial Allocate support
#4651 opened Jun 17, 2025 by wolfcomos Loading…
@wolfcomos
54
Adding a new benchmark for softmax fwd to track performance for large inner dimensions
#4650 opened Jun 17, 2025 by protonu • Draft
4
[RFC] Add Cutlass MXFP8 Grouped Gemm to nvfuser_direct python bindings Cutlass Matmuls Thunder-Inference-Demo
#4649 opened Jun 17, 2025 by rdspring1 Loading…
3
Option to run benchmarks with nsys
#4648 opened Jun 16, 2025 by Priya2698 Loading…
5
Implement Warp-Specialized Ping-Pong Matmul Matmuls
#4646 opened Jun 16, 2025 by rdspring1 • Draft
2
Fix condition for smem limit in kernel executor
#4644 opened Jun 16, 2025 by jacobhinkle • Draft
5
ws tma normalization dynamic shape
#4638 opened Jun 13, 2025 by liqiangxl • Draft
2
Add fused Embedding and RMSNorm benchmarks Python Benchmarks
#4637 opened Jun 13, 2025 by IvanYashchuk Loading…
11
Add embedding_indexing benchmark and Llama 4 Maverick configuration Python Benchmarks
#4636 opened Jun 13, 2025 by IvanYashchuk Loading…
1
Use CUPTI-python in lieu of torch.profiler
#4614 opened Jun 11, 2025 by Priya2698 Loading…
24
[WIP] Always do CGA split in persistent Hopper matmul
#4610 opened Jun 10, 2025 by jacobhinkle • Draft
6
[WIP] Inline stmatrix with TMA store
#4609 opened Jun 10, 2025 by jacobhinkle • Draft
2
7
auto select between warp specialized and multi-wave approaches
#4603 opened Jun 9, 2025 by liqiangxl Loading…
4
[DO NOT REVIEW] Adding index_shuffling
#4588 opened Jun 6, 2025 by jjsjann123 • Draft
2
Privatize squeeze in addition to upcast
#4583 opened Jun 5, 2025 by protonu Loading…
9
LLVM lowering
#4581 opened Jun 5, 2025 by wolfcomos Loading…
94
Add option to insert resharding after
#4574 opened Jun 4, 2025 by samnordmann Loading…
5
ProTip! Type g i on any issue or pull request to go back to the issue listing page.