Add AMD GPU support for DeepSeek-V3.2-Exp (MI300X/MI325X/MI355X + AITER)#279
Add AMD GPU support for DeepSeek-V3.2-Exp (MI300X/MI325X/MI355X + AITER)#279GoldenGrapeGentleman wants to merge 1 commit intovllm-project:mainfrom
Conversation
- Add Step 1: uv-based vLLM ROCm installation - Add Step 2: vLLM server startup with AITER and AITER_MOE enabled - Use --block-size 1 required for ROCm Sparse MLA - Set --kv-cache-dtype bfloat16 for better latency - Set --max-num-batched-tokens 32768 for throughput - Add Step 3: benchmark script Tested and verified on 8x AMD MI300X and 8x AMD MI355X GPUs with AITER. Signed-off-by: Yuan Yue <yueyuan@amd.com>
Summary of ChangesHello, 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 significantly enhances the DeepSeek-V3.2-Exp model's accessibility by integrating official support for AMD MI300X/MI325X/MI355X GPUs. It provides users with a clear, verified pathway to deploy and benchmark the model on AMD hardware, addressing a critical compatibility gap and expanding the model's operational environment. Highlights
🧠 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. Changelog
Activity
Using Gemini Code AssistThe 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
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 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
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request adds documentation for running DeepSeek-V3.2-Exp on AMD GPUs. The new section in DeepSeek-V3_2-Exp.md provides clear instructions for installation, server launch, and benchmarking. My review includes a suggestion to improve the conciseness of the introductory text for better readability.
| Recommended approaches by hardware type are: | ||
|
|
||
| MI300X/MI325X/MI355X | ||
|
|
||
| Please follow the steps here to install and run DeepSeek-V3.2-Exp models on AMD MI300X/MI325X/MI355X GPU. |
There was a problem hiding this comment.
The introductory text for the AMD GPU support section is a bit repetitive and could be more concise. Consider rephrasing to improve readability.
| Recommended approaches by hardware type are: | |
| MI300X/MI325X/MI355X | |
| Please follow the steps here to install and run DeepSeek-V3.2-Exp models on AMD MI300X/MI325X/MI355X GPU. | |
| This section provides instructions for running DeepSeek-V3.2-Exp on AMD MI300X, MI325X, and MI355X GPUs. |
1 participant
Summary
This PR adds AMD GPU support for DeepSeek-V3.2-Exp on MI300X/MI325X/MI355X GPUs.
Changes
--block-size 1required for ROCm Sparse MLA attention--kv-cache-dtype bfloat16for better latency on short requests--max-num-batched-tokens 32768for throughput optimizationHardware Tested
Related
Closes the AMD GPU support gap originally started in #143.