ZhangTao1999/AIQuantumCompiler

MCP Gate Optimize

A Model Context Protocol (MCP) server for quantum gate optimization in neutral atom systems.

Features

Pulse-Level Optimization (src/gate_optimize/pulse/)

• CZ Gate Optimization: GRAPE algorithm implementation for controlled-Z gate optimization
• X Gate Optimization: Robust GRAPE with Fourier coefficients for X gate optimization

Circuit-Level Optimization (src/gate_optimize/circuit/)

• Deep Q-Network (DQN): Reinforcement learning for circuit optimization
• Proximal Policy Optimization (PPO): Advanced RL algorithm for gate sequence optimization
• Environment Simulation: Custom RL environment for quantum circuit optimization

qec for any quantum code (src/gate_optimize/qec/)

• Quantum error correction: run decoder for the quantum code

Project Structure

gate_optimize/
├── src/gate_optimize/
│   ├── pulse/           # Hardware pulse optimization
│   ├── circuit/         # ML circuit optimization  
│   ├── qec/             # error correction code 
│   └── server.py        # MCP server implementation
├── model/               # Trained models and evaluation data
│   ├── plots/          # Training results and visualizations
│   ├── eval/           # Evaluation parameters and benchmarks
│   └── tests/          # Test configurations and datasets
└── pyproject.toml      # Project dependencies

MCP Tools

• optimize_cz_gate: Optimize CZ gate using GRAPE algorithm
• optimize_x_gate: Optimize X gate using GRAPE with Fourier coefficients

Installation

uv sync

Usage

Start the MCP server to access quantum gate optimization tools for neutral atom quantum computing research.

Command

Circuit generation

Some simple test

• Generate circuit from the GHZ/Steane code model/eval/7bit_params.json python -m gate_optimize.circuit.minimal_runner

• Optimize the CZ gate python -m gate_optimize.pulse.optimize_cz

• Optimize the X gate python -m gate_optimize.pulse.optimize_x