mcp-quantum/CONFIGURATION.md

MCP Quantum Server - Example Configurations

This directory contains example configurations for different use cases.

Configuration Files

Environment Configuration (.env)

# Server Configuration
SERVER_PORT=3000
SERVER_HOST=localhost
NODE_ENV=development

# MCP Configuration
MCP_SERVER_NAME=cuda-quantum-mcp
MCP_SERVER_VERSION=1.0.0

# CUDA Quantum Configuration
CUDAQ_PYTHON_PATH=/usr/local/bin/python3
CUDAQ_DEFAULT_TARGET=qpp-cpu
CUDAQ_LOG_LEVEL=info

# GPU Configuration  
CUDA_VISIBLE_DEVICES=0
NVIDIA_VISIBLE_DEVICES=all
CUDAQ_ENABLE_GPU=true

# Hardware Provider API Keys
QUANTUM_MACHINES_API_KEY=your_api_key_here
IONQ_API_KEY=your_api_key_here
QUANTINUUM_API_KEY=your_api_key_here

# Performance
MAX_CONCURRENT_JOBS=10
QUANTUM_CIRCUIT_TIMEOUT=30000
MAX_QUBITS=32
MAX_SHOTS=100000

Claude Desktop Integration

Add to your Claude Desktop configuration file:

{
  "mcpServers": {
    "cuda-quantum": {
      "command": "node",
      "args": ["/path/to/mcp-quantum/dist/index.js"],
      "env": {
        "CUDAQ_PYTHON_PATH": "/usr/local/bin/python3",
        "CUDAQ_DEFAULT_TARGET": "qpp-cpu",
        "LOG_LEVEL": "info"
      }
    }
  }
}

VS Code Launch Configuration

{
  "version": "0.2.0",
  "configurations": [
    {
      "name": "Debug MCP Server",
      "type": "node",
      "request": "launch",
      "program": "${workspaceFolder}/dist/index.js",
      "env": {
        "NODE_ENV": "development",
        "LOG_LEVEL": "debug",
        "CUDAQ_PYTHON_PATH": "/usr/local/bin/python3"
      },
      "console": "integratedTerminal",
      "sourceMaps": true,
      "restart": true,
      "runtimeArgs": ["--enable-source-maps"]
    }
  ]
}

Deployment Configurations

Docker Compose

version: '3.8'
services:
  mcp-quantum:
    build: .
    environment:
      - NODE_ENV=production
      - CUDAQ_DEFAULT_TARGET=qpp-gpu
      - LOG_LEVEL=info
      - CUDA_VISIBLE_DEVICES=0
    volumes:
      - ./logs:/app/logs
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]

Kubernetes Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  name: mcp-quantum-server
spec:
  replicas: 1
  selector:
    matchLabels:
      app: mcp-quantum-server
  template:
    metadata:
      labels:
        app: mcp-quantum-server
    spec:
      containers:
      - name: mcp-quantum
        image: mcp-quantum:latest
        env:
        - name: NODE_ENV
          value: "production"
        - name: CUDAQ_DEFAULT_TARGET
          value: "qpp-gpu"
        - name: LOG_LEVEL
          value: "info"
        resources:
          limits:
            nvidia.com/gpu: 1
          requests:
            memory: "2Gi"
            cpu: "1000m"

Hardware-Specific Configurations

IonQ Configuration

# IonQ specific settings
IONQ_API_KEY=your_ionq_api_key
IONQ_BASE_URL=https://api.ionq.co/v0.3
IONQ_DEFAULT_BACKEND=simulator
IONQ_MAX_SHOTS=10000

Quantinuum Configuration

# Quantinuum specific settings
QUANTINUUM_API_KEY=your_quantinuum_api_key
QUANTINUUM_BASE_URL=https://api.quantinuum.com/v1
QUANTINUUM_DEFAULT_DEVICE=H1-1E
QUANTINUUM_MAX_SHOTS=10000

Multi-GPU Configuration

# Multi-GPU setup
CUDA_VISIBLE_DEVICES=0,1,2,3
NVIDIA_VISIBLE_DEVICES=all
CUDAQ_GPU_MEMORY_FRACTION=0.8
CUDAQ_ENABLE_MULTI_GPU=true

Development Configurations

Testing Configuration

NODE_ENV=test
LOG_LEVEL=error
CUDAQ_DEFAULT_TARGET=qpp-cpu
QUANTUM_CIRCUIT_TIMEOUT=5000
MAX_SHOTS=1000

Debug Configuration

NODE_ENV=development
LOG_LEVEL=debug
CUDAQ_LOG_LEVEL=debug
PYTHON_TIMEOUT=120000
ENABLE_TRACE_LOGGING=true