Skip to content

Issues: NVIDIA/numba-cuda

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

30 Open 11 Closed
30 Open 11 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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[FEA] NVIDIA nvvm defaults to fma=1 and wish there was a way to tell it fma=0 feature request New feature or request
#118 opened Jan 28, 2025 by marioroy
[FEA] Make device attributes queryable feature request New feature or request
#111 opened Jan 17, 2025 by din14970
1
[FEA] Make block, thread, and warp indices unsigned feature request New feature or request
#109 opened Jan 9, 2025 by gmarkall
2
[BUG] test_empty_slice_1d/test_empty_slice_2d fail with NumPy 2.2.1 bug Something isn't working
#108 opened Jan 9, 2025 by leofang
[BUG] Debug: grid groups appear to be floats, in cuda-gdb bug Something isn't working
#107 opened Jan 8, 2025 by gmarkall
[FEA] Support for caching kernels using NRT / add target context to kernels loaded from cache feature request New feature or request
#105 opened Jan 8, 2025 by gmarkall
[FEA] Support for Graph APIs feature request New feature or request
#102 opened Jan 2, 2025 by gmarkall
[BUG] LinkableCode not working with CtypesLinker bug Something isn't working
#101 opened Jan 2, 2025 by gmarkall
[FEA] Add overloads generated by specialization to the current dispatcher feature request New feature or request
#100 opened Dec 31, 2024 by gmarkall
[FEA] support for nextafter feature request New feature or request
#99 opened Dec 31, 2024 by gmarkall
[BUG] Large overhead when launching kernel with torch tensors bug Something isn't working
#98 opened Dec 30, 2024 by gmarkall
[BUG] Kernel fixup does not adjust debug metadata arguments bug Something isn't working
#95 opened Dec 19, 2024 by gmarkall
1
[FEA] Provide a way to set the maximum dynamic shared memory size feature request New feature or request
#94 opened Dec 19, 2024 by Gui-Yom
6
[DOC] Configure doc publishing to disambiguate latest and released versions doc Improvements or additions to documentation
#87 opened Dec 9, 2024 by jakirkham
[drive-by-feedback] Packaging quirks bug Something isn't working
#85 opened Dec 7, 2024 by webknjaz
2
[BUG] Wheels publishing broken bug Something isn't working
#75 opened Nov 29, 2024 by gmarkall
2
[BUG] Advice for installing CUDA toolkit in error message is wrong bug Something isn't working
#69 opened Nov 21, 2024 by gmarkall
[FEA] Support automatic generation of link list for external code feature request New feature or request
#67 opened Nov 6, 2024 by gmarkall
4
[FEA] Support for loading NVRTC from a wheel feature request New feature or request
#66 opened Oct 29, 2024 by gmarkall nvmath-python support
1
[FEA] Support for loading NVVM from a wheel feature request New feature or request
#65 opened Oct 29, 2024 by gmarkall nvmath-python support
1
[BUG] Use of uninitialized value triggered in NVVM bug Something isn't working
#64 opened Oct 28, 2024 by gmarkall
[DOC] lto kwarg for the @cuda.jit decorator is undocumented doc Improvements or additions to documentation
#61 opened Oct 22, 2024 by gmarkall v0.5.0
[CI] Refactor CI matrix improvement Improves an existing functionality
#57 opened Oct 21, 2024 by gmarkall
[FEA] Dim3 should act more like a NamedTuple feature request New feature or request
#55 opened Oct 18, 2024 by gmarkall v0.5.0
1
[FEA] Support adding additional compile flags in nvrtc.compile feature request New feature or request
#46 opened Aug 23, 2024 by isVoid
ProTip! no:milestone will show everything without a milestone.