Skip to content

Pull requests: NVIDIA/cutlass

New pull request New
50 Open 702 Closed
50 Open 702 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

fix(conv int64 addressing): fixed large shape conv illegal memory access
#2651 opened Sep 18, 2025 by oyanghd Loading…
1
Fix incorrect description about the value of e2m1 in comment
#2646 opened Sep 17, 2025 by waveform Loading…
is_unsigned_v is only available if the code is compiled with C++17 or later:
#2621 opened Sep 5, 2025 by FurkanGozukara Loading…
2
fix typos
#2610 opened Sep 3, 2025 by co63oc Loading…
2
Fix bfloat16 epsilon
#2607 opened Sep 2, 2025 by gengar-maker Loading…
fit argument bug in sgemm examples.
#2605 opened Sep 2, 2025 by qmpzzpmq Loading…
Tiny fix tutorial text
#2594 opened Aug 26, 2025 by fzyzcjy Loading…
fix: missing argparse argument static_shape for cutedsl ampere sgemm example
#2568 opened Aug 14, 2025 by a-r-r-o-w Loading…
2
Fix arch guards in a few examples
#2567 opened Aug 13, 2025 by alihassanijr Loading…
1
7
Fix typo in tv layout stride in elementwise_add inactive-30d
#2564 opened Aug 12, 2025 by henrylhtsang Loading…
1
Add missing CUDA_ARCH guard for __nanosleep in example inactive-30d
#2558 opened Aug 11, 2025 by Flamefire Loading…
10
Liberate runtime check in example 13
#2554 opened Aug 7, 2025 by FabianSchuetze Loading…
2
Fix typo in cute.nvgpu.warpgroup.mma doc inactive-30d
#2548 opened Aug 7, 2025 by lucifer1004 Loading…
1
[DOC] Remove unrendering latex inactive-30d
#2527 opened Jul 31, 2025 by skailasa Loading…
4
fix: too early producer commit in examples/77_blackwell_fmha/kernel/sm100_fmha_mla_tma_warpspecialized.hpp
#2502 opened Jul 24, 2025 by soundOfDestiny Loading…
5
Mixed Precision Grouped Gemm with zero points and GPT-Q semantics closes #2261 inactive-30d
#2457 opened Jul 11, 2025 by ankutalev Loading…
1
7
Make unittest compilation faster inactive-30d
#2402 opened Jun 14, 2025 by andralex Loading…
10
Add SM80/89 blockwise scaling kernel, support FP8 block/groupwise on Ada, INT8 on Ampere
#2328 opened May 24, 2025 by solrex Loading…
8
Fix CMake build error and -Wdeprecated-literal-operator warnings
#2305 opened May 15, 2025 by BwL1289 Loading…
37
Support N={48, 80, 96, 112, ...} for SM100 EpilogueTileAuto inactive-30d inactive-90d
#2269 opened Apr 29, 2025 by Algy Loading…
4
Limit the number of SMs (sm_count) to user-provided value during profiling. inactive-30d inactive-90d
#2257 opened Apr 22, 2025 by manishucsd Loading…
5
Adding Blackwell support for distributed GEMM.
#2179 opened Mar 16, 2025 by whatdhack Loading…
5
fix: the bug of example_simt_canonical inactive-30d inactive-90d
#2160 opened Mar 10, 2025 by fengxiaoshuai Loading…
2
Request to remove annoyingly high number of warnings inactive-30d inactive-90d
#2141 opened Feb 26, 2025 by manishucsd Loading…
2
Allow cluster sizes across m,n,k to be reported in cutlass profiler inactive-30d inactive-90d
#2078 opened Feb 4, 2025 by mandroid6 Loading…
6
ProTip! Adding no:label will show everything without a label.