Improve profiler implementation and add comprehensive tests

[jainapurva]

No description provided.

[jainapurva]

Stack from ghstack (oldest at bottom):

• Memory profiler for cuda #1993
• -> Improve profiler implementation and add comprehensive tests #1992
• Add profiler and Perfetto UI link #1984

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1992

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 10 Pending

As of commit 668f776 with merge base 42e1345 ():

NEW FAILURES - The following jobs have failed:

• .github/workflows/float8nocompile_test.yaml (gh)
• Code Analysis with Ruff / build (3.9) (gh)
  benchmarks/microbenchmarks/benchmark_inference.py:17:8: F401 [*] uuid imported but unused

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[Copilot]

Pull Request Overview

This pull request improves the profiler implementation and adds comprehensive tests for profiling functionality.

• Refactors the profiler code to use a cached device variable and a warm-up phase for more consistent measurements.
• Adjusts profiler settings to reduce overhead by disabling stack traces and memory profiling.
• Introduces new tests to verify basic, CUDA, and memory profiling outputs.

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
benchmarks/microbenchmarks/utils.py	Updates profiler configuration and refactors device check.
benchmarks/microbenchmarks/test/test_benchmark_profiler.py	Adds comprehensive tests for profiler and memory profiling.