Leo Fang comments

Results 1175 comments of


                                            Leo Fang

feature request : implementation of `numpy.linalg.eig` equivalent

xref for `eigvals`: https://github.com/cupy/cupy/issues/6359

Possibly mistaken usage of the `cupyx.jit.thrust.device` execution policy in the JIT context?

Sorry, apparently I only typed the note without leaving the actual question... the question is: did you see issues when using the `seq` policy during development, thus deciding to switch...

Revisit: Using upstream Thrust complex headers and drop the vendored ones

cc: @jrhemstad for vis

Do not use plain thrust namespace in complex clone

/test mini

Do not use plain thrust namespace in complex clone

There are build- and run- time compilation failures regarding missing `` include. For run-time ones it's clear, since NVRTC doesn't have it, and we weren't expected to add it without...

Do not use plain thrust namespace in complex clone

> We probably need to bump the CCCL submodule so that [NVIDIA/cccl#1081](https://github.com/NVIDIA/cccl/pull/1081) is included (but I thought it already was?). Ah I see, the submodule was pinned around last September,...

Do not use plain thrust namespace in complex clone

Note: This is included as part of #8412.

Splayed layout support

Sorry for lack of response. Is this still an issue? It seems we have been able to build cuda-python on conda-forge?

Splayed layout support

@jakirkham I am sorry but I don't get it. In any case we need a way to specify where the CUDA headers are, and setting `CUDA_HOME` is the right solution...

Splayed layout support

I still don't understand what specific changes are needed to support splayed layout. @jakirkham this issue might have been outdated since the earlier (CTK 12.0) efforts? Right now, all headers...