OptimAI Network Node Setup Guide

Complete guide for setting up and running OptimAI Network edge nodes for AI data processing.

Overview

OptimAI Network is a decentralized AI infrastructure network that enables edge devices to contribute to AI model training, inference, and data processing. Node operators earn rewards based on compute contribution.

System Requirements

Minimum Requirements

• CPU: 4 cores (x86_64 or ARM64)
• RAM: 8 GB
• Storage: 50 GB SSD
• Network: 10 Mbps stable connection
• OS: Ubuntu 20.04/22.04, Debian 11/12, or compatible Linux

Recommended

• CPU: 8+ cores (Intel i7/i9 or AMD Ryzen 7/9)
• RAM: 16-32 GB DDR4
• Storage: 200 GB NVMe SSD
• GPU: NVIDIA GTX 1060+ or RTX series (optional but recommended)
• Network: 100 Mbps+ with low latency

GPU Acceleration (Optional)

• NVIDIA: CUDA 11.8+ compatible GPU
• AMD: ROCm support (limited)
• VRAM: 4 GB minimum, 8 GB+ recommended

Installation

Step 1: System Preparation

# Update system
sudo apt update && sudo apt upgrade -y

# Install dependencies
sudo apt install -y curl wget git python3 python3-pip docker.io docker-compose

# Add user to docker group
sudo usermod -aG docker $USER
newgrp docker

Step 2: Install NVIDIA Drivers (if using GPU)

# Add NVIDIA package repositories
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list

# Install nvidia-docker2
sudo apt update
sudo apt install -y nvidia-docker2

# Restart Docker
sudo systemctl restart docker

# Test GPU access
docker run --rm --gpus all nvidia/cuda:11.8.0-base-ubuntu22.04 nvidia-smi

Step 3: Install OptimAI CLI

# Download OptimAI CLI
curl -sSL https://install.optimai.network/optimai-cli.sh | bash

# Or manual installation
wget https://github.com/optimai-network/optimai-cli/releases/latest/download/optimai-linux-amd64
chmod +x optimai-linux-amd64
sudo mv optimai-linux-amd64 /usr/local/bin/optimai

# Verify installation
optimai --version

Step 4: Authenticate and Register

# Login with your OptimAI account
optimai login

# Or use API key
optimai login --api-key YOUR_API_KEY

# Register node
optimai node register --name "your-node-name"

Step 5: Configure Node

mkdir -p ~/.optimai

cat > ~/.optimai/config.yaml << 'EOF'
node:
  name: "carly-optimai-node"
  region: "asia-southeast"
  
compute:
  cpu:
    enabled: true
    cores: 4
    priority: "normal"
  gpu:
    enabled: true
    device: "nvidia"
    memory_limit: "6GB"
  memory:
    limit: "12GB"
  storage:
    cache_size: "100GB"
    
network:
  listen_port: 8080
  public_ip: "auto"
  max_bandwidth: "100Mbps"
  
workloads:
  inference: true
  training: true
  data_processing: true
  
rewards:
  address: "YOUR_ETHEREUM_ADDRESS"
  auto_claim: true
  
monitoring:
  enabled: true
  metrics_port: 9090
  log_level: "info"
EOF

Step 6: Start Node

Option A: Docker Compose (Recommended)

cd ~/.optimai

cat > docker-compose.yml << 'EOF'
version: '3.8'

services:
  optimai-node:
    image: optimai/network-node:latest
    container_name: optimai-node
    restart: unless-stopped
    runtime: nvidia  # Remove if no GPU
    ports:
      - "8080:8080"
      - "9090:9090"
    volumes:
      - ./config.yaml:/app/config.yaml
      - ./cache:/app/cache
      - ./models:/app/models
    environment:
      - OPTIMAI_CONFIG=/app/config.yaml
      - NVIDIA_VISIBLE_DEVICES=all
    deploy:
      resources:
        limits:
          cpus: '4'
          memory: 12G
        reservations:
          cpus: '2'
          memory: 4G
EOF

# Start node
docker-compose up -d

# View logs
docker-compose logs -f

Option B: Systemd Service

sudo tee /etc/systemd/system/optimai-node.service > /dev/null << 'EOF'
[Unit]
Description=OptimAI Network Node
After=network.target docker.service
Requires=docker.service

[Service]
Type=simple
User=$USER
WorkingDirectory=$HOME/.optimai
ExecStart=/usr/bin/docker-compose up
ExecStop=/usr/bin/docker-compose down
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target
EOF

sudo systemctl daemon-reload
sudo systemctl enable optimai-node
sudo systemctl start optimai-node

Monitoring

Check Node Status

# Node health
optimai node status

# Compute resources
optimai node resources

# Active workloads
optimai workloads list

# Rewards
optimai rewards status

View Logs

# Docker logs
docker logs -f optimai-node

# Systemd logs
sudo journalctl -u optimai-node -f

# Metrics
curl http://localhost:9090/metrics

Dashboard

Access node dashboard at: https://dashboard.optimai.network

Troubleshooting

Issue: Node authentication fails

Symptoms: optimai login fails

Solutions:

1. Check API key:

   optimai login --api-key YOUR_KEY

2. Verify network:

   curl https://api.optimai.network/health

3. Reset credentials:

   rm ~/.optimai/credentials
   optimai login

Issue: GPU not detected

Symptoms: Node running but GPU not used

Solutions:

1. Check NVIDIA drivers:

   nvidia-smi

2. Verify nvidia-docker:

   docker run --rm --gpus all nvidia/cuda:11.8.0-base-ubuntu22.04 nvidia-smi

3. Update config:

   gpu:
     enabled: true
     device: "nvidia"

4. Restart node:

   docker-compose restart

Issue: Low task allocation

Symptoms: Node online but few tasks

Solutions:

1. Check compute availability:

   optimai node resources

2. Enable more workload types:

   workloads:
     inference: true
     training: true
     data_processing: true

3. Improve network latency

4. Upgrade hardware specs

Issue: Out of memory

Symptoms: Node crashes with OOM errors

Solutions:

1. Reduce memory limit:

   memory:
     limit: "8GB"

2. Limit concurrent tasks:

   compute:
     max_concurrent: 2

3. Add swap:

   sudo fallocate -l 8G /swapfile
   sudo chmod 600 /swapfile
   sudo mkswap /swapfile
   sudo swapon /swapfile

Issue: Model download fails

Symptoms: Cannot download AI models

Solutions:

1. Check storage:

   df -h

2. Clear model cache:

   rm -rf ~/.optimai/cache/models/*

3. Check network:

   ping huggingface.co

Issue: No rewards

Symptoms: Node running but no rewards

Solutions:

1. Verify reward address:

   optimai config get rewards.address

2. Check task completion rate:

   optimai workloads history

3. Verify minimum uptime (usually 24h)

4. Check dashboard for penalties

FAQ

Q: What AI models are supported? A: OptimAI supports various LLMs, vision models, and custom models. Check dashboard for current list.

Q: Can I run without GPU? A: Yes, but GPU significantly increases earning potential.

Q: How are rewards calculated? A: Based on compute contributed (FLOPS), task completion rate, and network demand.

Q: Is my data private? A: Yes, all processing is done locally. Only anonymized metrics are sent to the network.

Q: Can I limit resource usage? A: Yes, configure limits in config.yaml for CPU, GPU, memory, and bandwidth.

Performance Optimization

CPU Optimization

compute:
  cpu:
    cores: 6
    affinity: "0-5"  # Pin to specific cores
    governor: "performance"

GPU Optimization

gpu:
  enabled: true
  device: "nvidia"
  memory_limit: "10GB"
  power_limit: 250  # Watts
  clock_offset: 100

Network Optimization

network:
  max_bandwidth: "500Mbps"
  prefer_ipv6: false
  keep_alive: true

Resources

• Official Documentation
• OptimAI Dashboard
• Discord Community
• GitHub
• Hugging Face

License

MIT