Alpaka VecMem Interface, main branch (2025.05.01.) #959
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Meant as an example/guidance for Ryan Cross.
krasznaa
force-pushed
the
AlpakaVecmemOrganization-main-20250501
branch
from
fcb7faf to
fac284a
Compare
|
|
Thanks for this! I'll take a look today and look at integrating it back into the Alpaka algorithms, and see how I get on. |
|
PR #960 is now opened with this commit as the base. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is a proposal for how I'd imagine using Alpaka in the longer term. This was triggered by #958, whose code I didn't really like. But since I have very little experience with Alpaka, I wanted to try out for myself what sort of code could be written.
@CrossR, @StewMH, this is how I imagine that things could look like. 🤔
traccc::alpaka::queueis the same sort of thing that you tried to make as well Ryan. I just simplified it so that it would only expose the underlying Alpaka queue, and not try to be too smart with the native devices/queues.traccc::alpaka::vecmem_objectshas the interface that Ryan made, just with a (in my mind) much simpler implementation.You see, once you hide the choice of VecMem objects behind such a non-specific interface, you can just use simple pre-processor statements in the background. There's no need to mix pre-processor formalism with traits in this case.
Please check if this sort of a setup could be made to work with the existing algorithms. Since that I didn't really try...