[ARROW-4212] [Python] [CUDA] Creating a CUDA buffer from Numba device array should be easier

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.11.1
• Fix Version/s: 0.13.0
• Component/s: GPU, Python
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/20792
• Assignee: Pearu Peterson
• Reporter: Antoine Pitrou
• Created: 09/Jan/19 16:49
• Updated: 11/Jan/23 07:32
• Resolved: 24/Jan/19 13:22

Description

Currently, to create a CUDA buffer from a Numba device array, you have to write:

cuda.CudaBuffer.from_numba(device_arr.gpu_data)

It would be easier if you could just write:

cuda.CudaBuffer.from_numba(device_arr)

(ideally, any object exposing the CUDA Array Interface would work)

Issue Links

• links to: GitHub Pull Request #3439