cuda-python
cuda-python copied to clipboard
NVIDIA
CUDA Python: Performance meets Productivity
Creating this as a place for myself to collect information related to finding libnvvm.so Note that libnvvm.so is unusual: while most CTK .so files live under \/lib or \/lib64, it...
`cuda.core` can benefit from clean and detailed type hints. The typing annotation can guide users to understand the API design and expected workflow.
This is ready for at least a first look Add Mempool class as the first public MemoryResource implementation. It supports IPC. Add IPC buffer which is an internal buffer implementation...
For legacy library, it takes time to migrate from module dependent APIs to library dependent APIs. In the transition time, `cuLibraryGetModule` is a very useful API to allow switching from...
Follow-up of #487. Like CUDA streams, events today are also tied to a CUDA context, so it makes sense to offer the reverse link when an `Event` is created. This...
In a multi-GPU setting, right now it's not possible to memcpy directly between the two devices. A couple of checks would have to be done by cuda.core as part of...
When the Driver and Runtime bindings are generated, we redefine the types from headers: - cudaEGL.h - cudaGL.h - cudaVDPAU.h - cuda_egl_interop.h - cuda_gl_interop.h - cuda_vdpau_interop.h We do this so...
We've introduced typing (Python type hints) support on Day 1 of the project, because it helps users understand `cuda.core`'s API design logic. We should ensure our typing is done correctly...
In this CI run we hit a bizarre NVRTC not found error at test time. However, it should have been properly installed prior to test execution. Turns out that Powershell...