io.github.andylbrummer/neural-mcp
GPU-accelerated MCP server for neural network training, deep learning, and model experimentation
★ 1MITdevtools
Install
Config snippet generator goes here (5 client tabs)
README
# Math-Physics-ML MCP System
[](https://pypi.org/project/scicomp-math-mcp/)
[](https://pypi.org/project/scicomp-quantum-mcp/)
[](https://pypi.org/project/scicomp-molecular-mcp/)
[](https://pypi.org/project/scicomp-neural-mcp/)
[](https://andylbrummer.github.io/math-mcp/)
[](https://opensource.org/licenses/MIT)
GPU-accelerated [Model Context Protocol](https://modelcontextprotocol.io) servers for computational mathematics, physics simulations, and machine learning.
## 📚 Documentation
**[View Full Documentation →](https://andylbrummer.github.io/math-mcp/)**
| Guide | Description |
|-------|-------------|
| [Installation](https://andylbrummer.github.io/math-mcp/getting-started/installation) | Setup instructions for pip, uv, and uvx |
| [Configuration](https://andylbrummer.github.io/math-mcp/getting-started/configuration) | Claude Desktop & Claude Code setup |
| [Quick Start](https://andylbrummer.github.io/math-mcp/getting-started/quick-start) | Get running in 5 minutes |
| [API Reference](https://andylbrummer.github.io/math-mcp/api/overview) | Complete tool documentation |
| [Visual Demos](https://andylbrummer.github.io/math-mcp/demos/) | Interactive physics simulations |
## About
This system enables AI assistants to perform real scientific computing — from solving differential equations to running molecular dynamics simulations.
<table>
<tr>
<td align="center" width="50%">
<img src="docs/static/img/demos/double_slit.webp" alt="Double-Slit Interference" width="100%"/>
<br/><b>Quantum Wave Mechanics</b><br/>
<sub>Double-slit interference pattern from solving the time-dependent Schrödinger equation</sub>
</td>
<td align="center" width="50%">
<img src="docs/static/img/demos/galaxy_collision.webp" alt="Galaxy Collision" width="100%"/>
<br/><b>N-Body Dynamics</b><br/>
<sub>Galaxy merger simulation using gravitational N-body calculations</sub>
</td>
</tr>
<tr>
<td align="center" width="50%">
<img src="docs/static/img/demos/bragg_hexagonal.webp" alt="Bragg Scattering" width="100%"/>
<br/><b>Crystal Diffraction</b><br/>
<sub>Bragg scattering from a hexagonal (graphene-like) lattice</sub>
</td>
<td align="center" width="50%">
<img src="docs/static/img/demos/triple_slit.webp" alt="Triple-Slit" width="100%"/>
<br/><b>Multi-Slit Interference</b><br/>
<sub>Complex interference patterns from three coherent sources</sub>
</td>
</tr>
</table>
## Overview
This system provides **4 specialized MCP servers** that bring scientific computing capabilities to AI assistants like Claude:
| Server | Description | Tools |
|--------|-------------|-------|
| **Math MCP** | Symbolic algebra (SymPy) + numerical computing | 14 |
| **Quantum MCP** | Wave mechanics & Schrodinger simulations | 12 |
| **Molecular MCP** | Classical molecular dynamics | 15 |
| **Neural MCP** | Neural network training & evaluation | 16 |
**Key Features:**
- GPU acceleration with automatic CUDA detection (10-100x speedup)
- Async task support for long-running simulations
- Cross-MCP workflows via URI-based data sharing
- Progressive discovery for efficient tool exploration
## Quick Start
### Installation with uvx (Recommended)
Run any MCP server directly without installation:
```bash
# Run individual servers
uvx scicomp-math-mcp
uvx scicomp-quantum-mcp
uvx scicomp-molecular-mcp
uvx scicomp-neural-mcp
```
### Installation with pip/uv
```bash
# Install individual servers
pip install scicomp-math-mcp
pip install scicomp-quantum-mcp
pip install scicomp-molecular-mcp
pip install scicomp-neural-mcp
# Or install all at once
pip install scicomp-math-mcp scicomp-quantum-mcp scicomp-molecular-mcp scicomp-neural-mcp
# With GPU support (requires CUDA)
pip install scicomp-math-mcp[gpu] scicomp-quantum-mcp[gpu] scicomp-molecular-mcp[gpu] scicomp-neural-mcp[gpu]
```
## Configuration
### Claude Desktop
Add to your Claude Desktop configuration file:
**macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json`
**Windows**: `%APPDATA%\Claude\claude_desktop_config.json`
```json
{
"mcpServers": {
"math-mcp": {
"command": "uvx",
"args": ["scicomp-math-mcp"]
},
"quantum-mcp": {
"command": "uvx",
"args": ["scicomp-quantum-mcp"]
},
"molecular-mcp": {
"command": "uvx",
"args": ["scicomp-molecular-mcp"]
},
"neural-mcp": {
"command": "uvx",
"args": ["scicomp-neural-mcp"]
}
}
}
```
### Claude Code
Add to your project's `.mcp.json`:
```json
{
"mcpServers