Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

SYCL. Add dask support for distributed calculations with sycl #10812

Merged
merged 5 commits into from
Sep 21, 2024

Conversation

razdoburdin
Copy link
Contributor

Hi,
I continue adding of sycl support (#10806, #10701, etc).
This PR adds support for distributed calculations with sycl and add one simple test for basic verification.

Copy link

@RahulVadisetty91 RahulVadisetty91 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

image

The from_pystr_to_cstr function has overload annotations but lacks a complete implementation. Make sure that the function implementation is correctly handling both cases described in the overloads.

" managed by the distributed frameworks. use `device=cuda` or `device=gpu`"
" instead."
)
if device != "sycl:gpu":
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What happens if it's sycl GPU, is dask going to select the appropriate GPU?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, the appropriate GPUs would be selected

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it.

for (size_t i = 0; i < sync_ids.size(); i++) {
auto& this_hist = hist_[sync_ids[i]];
const GradientPairT* psrc = reinterpret_cast<const GradientPairT*>(this_hist.DataConst());
// std::copy(psrc, psrc + nbins, reduce_buffer.begin() + i * nbins);
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please remove the commented code.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

done

@razdoburdin
Copy link
Contributor Author

Hi @trivialfis ,
what is your opinion about this PR?

@trivialfis trivialfis merged commit d7599e0 into dmlc:master Sep 21, 2024
30 checks passed
@razdoburdin razdoburdin deleted the dev/sycl/dask_support branch September 23, 2024 13:45
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants