Support the CUDA Array Interface
This commit adds support for version 3 of the CUDA Array Interface (https://numba.readthedocs.io/en/latest/cuda/cuda_array_interface.html): - GPU Arrays export the interface so that they can be passed to other supporting libraries. - Function calls check for the presence of the interface and use the data pointer from it if found, so other libraries' arrays can be passed to PyCUDA-compiled kernels. Documentation, tests, and examples for the interface going in both directions are added.
Loading
Please register or sign in to comment