Skip to content

CONTRIBUTING.md

Latest commit

聽

History

History
59 lines (43 loc) 路 2.4 KB

CONTRIBUTING.md

File metadata and controls

59 lines (43 loc) 路 2.4 KB
Raw

MAX AI kernels contributor guide

Thank you for your interest in contributing to the MAX AI kernels! We鈥檙e excited to welcome external contributors to help build high-performance CPU and GPU kernels in Mojo.

Contributing to MAX AI kernels

The MAX AI kernels power key components of MAX and Mojo, and contributions here can have a meaningful impact across the AI ecosystem. Before jumping in, please take a moment to review our contribution guidelines and processes.

We welcome contributions across a wide range of functionality, but if you鈥檙e proposing a change that could affect the performance of the core kernel library, we ask that you first go through our proposal process.

Changes we accept

We welcome contributions across all hardware platforms and are particularly interested in Blackwell, Hopper, MI3xx, and other data center GPUs.

In particular, we welcome contributions of the following kernels:

  • Batched matrix multiplication (BMM)
  • Multi-head latent attention (MLA)
  • Mixture of experts (MOE)
  • Non-maximum suppression (NMS)
  • Grouped matrix multiplication
  • 2D convolutions
  • General matrix-vector multiply (GEMV) on Hopper

Proposals

If you鈥檙e interested in making a significant change (one that doesn鈥檛 fall into the list of changes we accept above), your first step is a written proposal. The proposal process:

  • Ensures feedback from the widest possible set of community members
  • Maintains an audit trail of important design decisions
  • Provides rationale for changes that impact long-term performance or architecture

To submit a proposal, create a GitHub Pull Request that adds a document to the modular/mojo/proposals directory.

Contributors are encouraged to react with a 馃憤 to proposal PRs if they support the high-level direction. Proposals are reviewed and assigned to MAX AI kernels leads. A proposal can be merged once the lead approves, all blocking issues are resolved, and relevant decisions are incorporated. If a lead chooses to defer or reject the proposal, they will explain the reasoning and close the PR. This process is inspired by successful practices in other open source communities.

Submitting pull requests

For details about how to submit a pull request, see the repo's primary contributing guide.

Thank you for your contributions! 鉂わ笍