[FEA] Refactor device/runtime APIs in `rmm._cuda.gpu`

[jakirkham]

Is your feature request related to a problem? Please describe.

Currently rmm._cuda.gpu contains both device and runtime APIs. Based on discussion ( https://github.com/rapidsai/cudf/pull/4943#pullrequestreview-395881420 ), it would be good to refactor these.

Describe the solution you'd like

Ideally these would be refactored into 2 modules so that users could pull APIs from either one as needed.

Describe alternatives you've considered

Suggestions welcome 🙂

Additional context

NA

[kkraus14]

The cudf._cuda module has been moved to rmm so transferring issue.

[jakirkham]

cc @galipremsagar

[harrism]

Currently rmm._cuda.gpu contains both device and runtime APIs.

Do you mean (CUDA) driver and runtime APIs?

[github-actions[bot]]

This issue has been marked rotten due to no recent activity in the past 90d. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

[github-actions[bot]]

This issue has been marked stale due to no recent activity in the past 30d. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be marked rotten if there is no activity in the next 60d.

[harrism]

@jakirkham @kkraus14 still relevant?

[kkraus14]

Yes but low priority.

[github-actions[bot]]

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

[github-actions[bot]]

This issue has been labeled inactive-90d due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.

[vyasr]

I'm going to close this as wontfix. Since this issue was originally created, the advent of cuda-python means that users shouldn't be getting this functionality from rmm anyway, and rmm's implementations have been stripped down thanks to calling cuda-python APIs. At this point rmm_cuda.gpu is a fairly small, internal-only, largely static module. We can reconsider if it needs to expand or move in any way in the future.