BeClaude

metal-kernel

New
97.3kSmitheryDevelopmentby pytorch

Write Metal/MPS kernels for PyTorch operators. Use when adding MPS device support to operators, implementing Metal shaders, or porting CUDA kernels to Apple Silicon. Covers native_functions.yaml dispatch, host-side operators, and Metal kernel implementation.

First seen 5/22/2026

Install & Usage

1
Open your MCP config
~/.claude.json
2
Add the server config

Add the configuration to "mcpServers": { "metal-kernel": { "command": "...", "args": [] } }

3
Restart Claude Code
/mcp
View source on GitHub
codingai-&-ml

Security Audits

LicenseUnknownSourceWarnRepositoryPass

Frequently Asked Questions

What is metal-kernel?

Write Metal/MPS kernels for PyTorch operators. Use when adding MPS device support to operators, implementing Metal shaders, or porting CUDA kernels to Apple Silicon. Covers native_functions.yaml dispatch, host-side operators, and Metal kernel implementation.

How to install metal-kernel?

To install metal-kernel: open your mcp config (~/.claude.json), then add the config to "mcpServers": { "metal-kernel": { "command": "...", "args": [] } }. Finally, /mcp in Claude Code.

What is metal-kernel best for?

metal-kernel is a mcp categorized under Development. It is designed for: coding, ai-&-ml. Created by pytorch.