[Fix] Only add SM90 and SM100 to check for auto-enabling TRT Allreduce Fusion

[b8zhong]

Motivation

Fix #15650

Modifications

Accidentally did not check Flashinfer repo. It is not compatible with SM120, so we shouldn't enable it here.

Source from Flashinfer (https://github.com/flashinfer-ai/flashinfer/blob/main/flashinfer/jit/comm.py)

  def gen_trtllm_comm_module() -> JitSpec:
      nvcc_flags = current_compilation_context.get_nvcc_flags_list(
          supported_major_versions=[9, 10]
      )

Accuracy Tests

N/A

Benchmarking and Profiling

N/A

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

[gemini-code-assist]

Summary of Changes

Hello @b8zhong, 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!

This pull request addresses a compatibility issue by adjusting the logic for automatically enabling TensorRT (TRT) Allreduce Fusion. Previously, the fusion could be activated on GPU architectures, such as SM120, that do not support it. The modification restricts the auto-enabling mechanism to only SM90 and SM100 architectures, ensuring that the feature is only utilized on compatible hardware and preventing potential runtime errors or incorrect behavior.

Highlights

• TRT Allreduce Fusion Scope: The auto-enabling condition for TensorRT (TRT) Allreduce Fusion has been refined to exclusively target SM90 (Hopper) and SM100 (Blackwell) GPU architectures.
• Compatibility Fix: This change resolves a compatibility issue where TRT Allreduce Fusion was inadvertently enabled for unsupported architectures, specifically SM120, which is not compatible with this feature.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request 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. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

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. ↩

[b8zhong]

/tag-and-rerun-ci

[gemini-code-assist]

Code Review

This pull request restricts the auto-enabling of TRT Allreduce Fusion to SM90 and SM100 architectures, removing support for SM120 which is not compatible. The change aligns with the provided context from Flashinfer's source code. I've added one comment regarding an outdated code comment that should be updated for clarity and maintainability.