Add kernel profiling support
Open
CPU
enhancement
- Dominant language
- C++
- Stars
- 54
- Forks
- 16
- PR merge metrics
- No merged PRs in 30d
Description
For a fused kernel, we should provide detailed profiling result to show the elapsed time for each fused components, like matmul, bias_add and relu. This can help performance analysis.
Contributor guide
Research direction
No files, tests, or entry points are named. Start by locating the fused-kernel execution and profiling paths in the repository, then determine how elapsed time can be attributed to components such as matmul, bias_add, and relu. Done means a fused kernel exposes detailed per-component timing for performance analysis.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- compilers, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100