feat(gpu): add Ascend NPU telemetry backend

[VincentCheungKokomo]

Summary

• add an Ascend NPU backend with DCMI-first telemetry and bounded npu-smi fallback
• expose NPU utilization, memory, bandwidth, temperature, and power through the existing GPU collector while respecting backend filters
• avoid native runtime activation in helper processes and document the new NPU controls

Test plan

• cargo test -p probing-gpu --features cuda
• cargo test -p probing-gpu --no-default-features
• cargo clippy -p probing-gpu --all-targets --features cuda -- -D warnings
• PROBING=0 .venv/bin/python -m pytest tests/regression/core/test_init.py -q
• validate DCMI ABI and telemetry values on Ascend hardware

Made with Cursor

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.