[Do Not Merge Now] Integrate new deepgemm

[yewentao256]

Purpose

DeepGemm is updating to v2.0 deepseek-ai/DeepGEMM#112

All of the API interfaces change and we migrate to their new version.

Test

Please see previous pr in #19820

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

[gemini-code-assist]

Summary of Changes

Hello @yewentao256, 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 is a significant migration effort to integrate the breaking API changes introduced in DeepGEMM v2.0 into the vLLM codebase. The primary goal is to maintain compatibility with the latest version of the DeepGEMM library, which is crucial for optimized FP8 GEMM operations. The changes involve updating function calls across various modules, adjusting import paths for DeepGEMM utilities, and temporarily incorporating some DeepGEMM helper functions directly into vLLM until they are officially re-exposed in DeepGEMM's public API.

Highlights

• DeepGEMM API Migration: Updated all usages of the DeepGEMM library to its new v2.0 API, reflecting significant interface changes. This includes renaming core GEMM functions (e.g., gemm_fp8_fp8_bf16_nt to fp8_gemm_nt) and grouped GEMM functions (e.g., m_grouped_gemm_fp8_fp8_bf16_nt_masked to fp8_m_grouped_gemm_nt_masked).
• Utility Function Relocation: Adjusted imports for DeepGEMM utility functions (like per_token_cast_to_fp8 and per_block_cast_to_fp8) to use their new locations within the deep_gemm.utils.math module. Additionally, some previously DeepGEMM-provided utilities (ceil_div, get_tma_aligned_size, get_col_major_tma_aligned_tensor) were copied locally into vLLM's fp8_utils.py as a temporary measure, with a TODO to remove them once DeepGEMM re-exposes them publicly.
• Benchmark and Test Updates: Modified existing benchmarks (benchmark_fp8_block_dense_gemm.py, benchmark_moe.py) and unit tests (test_deepep_deepgemm_moe.py, test_block_fp8.py) to align with the new DeepGEMM API. This ensures continued correctness and performance evaluation of FP8 operations within vLLM.

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 is currently in preview and 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 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 integrates the new DeepGEMM library, updating the API interfaces and migrating to the new version. The changes include modifications to benchmark scripts and test files to accommodate the new DeepGEMM API. The code has been reviewed and suggestions have been provided to improve code clarity and efficiency.