perf: cache get_compute_capability

[cyx-6]

📌 Description

Cache compute capability to reduce repeat cpu overload mentioned in #1425 (comment)

Results reproduced from https://gist.github.com/yzh119/d9bf2abbb667abcbb806979f4bbea633 :

before the fix
w/o CUDAGraph 0.0054492950439453125
w/  CUDAGraph 0.002916574478149414
after the fix
w/o CUDAGraph 0.0038330554962158203
w/  CUDAGraph 0.0030286312103271484

🔍 Related Issues

#1425

🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete.

✅ Pre-commit Checks

• I have installed pre-commit by running pip install pre-commit (or used your preferred method).
• I have installed the hooks with pre-commit install.
• I have run the hooks manually with pre-commit run --all-files and fixed any reported issues.

If you are unsure about how to set up pre-commit, see the pre-commit documentation.

🧪 Tests

• Tests have been added or updated as needed.
• All tests are passing (unittest, etc.).

Reviewer Notes

[gemini-code-assist]

Summary of Changes

Hello @cyx-6, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

I've implemented a performance optimization to cache the result of the get_compute_capability function. This change aims to reduce repeated CPU overload by ensuring that the compute capability is calculated only once for a given device and then retrieved from the cache for all subsequent calls, thereby improving efficiency.

Highlights

• Performance Optimization: I have applied the @functools.cache decorator to the get_compute_capability function. This change will ensure that the result of this function is cached after its first execution, preventing redundant computations and reducing CPU overhead on subsequent calls with the same arguments.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments or fill out our survey to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces a performance optimization by caching the result of get_compute_capability. This is a good change as it avoids re-computing the device's compute capability, which is a constant value during execution. However, I've identified a potential compatibility issue with the chosen caching decorator that could break support for Python 3.8, which the project seems to support based on other syntax used. My review includes a suggestion to fix this.

[gemini-code-assist]

@functools.cache was introduced in Python 3.9. The codebase uses features from Python 3.8 (like positional-only arguments in function definitions), which suggests that Python 3.8 compatibility might be intended. To ensure compatibility with Python 3.8, please use @functools.lru_cache(maxsize=None), which provides the same unbounded caching functionality and is available in Python 3.8. The codebase seems to use both @functools.cache and @functools.lru_cache, which could lead to inconsistencies.

Suggested change

	@functools.cache
	@functools.lru_cache(maxsize=None)

[yzh119]

cc @xiaoqi35, this PR should fix #1425