Use also a SM120 GPU in light CI for CCCL.C

[bernhardmgruber]

#8922 broke the build because the CI did not catch a CUB change affecting CCCL.C. CCCL.C only runs the light PR CI when CUB changes, which only tests on SM75. This PR adds a second run with SM120.

Thanks to @alliepiper for the analysis!

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: a6362c9e-fcf6-48f5-9e7c-0547c394f303

📥 Commits

Reviewing files that changed from the base of the PR and between faa9f1a and 9570b5a.

📒 Files selected for processing (1)

• ci/matrix.yaml

---

📝 Walkthrough

Summary by CodeRabbit

• Chores
  • Extended CI/CD test coverage for improved quality assurance across additional hardware configurations.

Walkthrough

The pull_request_lite CI workflow matrix is extended with a new test configuration for cccl_c_parallel targeting rtxpro6000 GPU on CTK 13.X with gcc13, adding GPU variant coverage to pull request validation.

Changes

CI Matrix Configuration

Layer / File(s)	Summary
Pull Request Lite Matrix - RTXPro6000 Target ci/matrix.yaml	New matrix entry adds rtxpro6000 GPU testing for cccl_c_parallel (CTK 13.X, gcc13, sm: gpu) to pull_request_lite workflow alongside existing rtx2080 coverage.

Suggested reviewers

• elstehle
• davebayer

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

😬 CI Workflow Results

🟥 Finished in 1h 57m: Pass: 99%/500 | Total: 5d 03h | Max: 59m 14s | Hits: 99%/622174

See results here.