DOC: Clarify requirements for running on GPU and supported configurations

[david-cortes-intel]

Description

Add a comprehensive description of proposed changes

List associated issue number(s) if exist(s): #6 (for example)

Documentation PR (if needed): #1340 (for example)

Benchmarks PR (if needed): IntelPython/scikit-learn_bench#155 (for example)

---

This PR:

• Updates the information about which configurations are supported on each distribution channel (latest conda-forge release has DPC and SPMD modules).
• Updates the list of supported python and sklearn versions to reflect the current situation in the main branch.
• Updates the description of GPU support logic.
• Updates the requirements on GPU software that is needed.
• Improves the examples by adding direct links to DPCTL docs, more code snippets, fixing typos, etc.

Checklist to comply with before moving PR from draft:

PR completeness and readability

• I have reviewed my changes thoroughly before submitting this pull request.
• I have updated the documentation to reflect the changes or created a separate PR with update and provided its number in the description, if necessary.
• Git commit message contains an appropriate signed-off-by string (see CONTRIBUTING.md for details).
• I have added a respective label(s) to PR if I have a permission for that.
• I have resolved any merge conflicts that might occur with the base branch.

Testing

• I have run it locally and tested the changes extensively.

Performance

Not applicable.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Flag	Coverage Δ
azure	77.96% <ø> (ø)
github	71.00% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

[syakov-intel]

LGTM

[ethanglaser]

Very nice work on this, the docs are much improved

[Alexsandruss]

Not entirely true: conda-forge linux build contains dpc and spmd_dpc libs starting from 2025 release. However, GPU support is not tested for it due to HW unavailability in conda-forge-feedstock CI.

[david-cortes-intel]

@Alexsandruss Thanks for the clarification.

But does the build from conda-forge differ from the build from intel-conda in some way that would affect GPU support? As I understand it, both of them are built with the exact same build-time dependencies (compiler, headers, TBB, MKL, MPI) in the exact same versions.

[Alexsandruss]

@Alexsandruss Thanks for the clarification.

But does the build from conda-forge differ from the build from intel-conda in some way that would affect GPU support? As I understand it, both of them are built with the exact same build-time dependencies (compiler, headers, TBB, MKL, MPI) in the exact same versions.

Toolchains are different: intel conda channel used internal, conda-forge - public.

[david-cortes-intel]

But then again: what differences are there in those toolchains that could affect GPU support, if they are at the same version numbers?

[maria-Petrova]

LGTM