How to Set Up Your First Decentralized AI Node (Step-by-Step)

What You'll Build

By the end of this guide, you'll have a fully operational decentralized AI compute node — connected to a network, accepting workloads, and earning token rewards. No prior Linux or AI experience required.

We'll cover: installing the operating system, setting up GPU drivers, configuring Docker for AI workloads, and registering on your chosen platform.

Prerequisites

• A computer with an NVIDIA GPU (RTX 3060 or better recommended)
• A USB drive (8GB+) for the installer
• A stable internet connection (50+ Mbps)
• A monitor and keyboard for initial setup (can go headless after)
• 30–60 minutes of uninterrupted time

Step 1: Install Ubuntu Server 24.04 LTS

We use Ubuntu Server (not Desktop) because it runs headless with minimal resource overhead — more power for AI workloads.

1. Download Ubuntu Server 24.04 LTS from ubuntu.com/download/server
2. Flash the ISO to your USB drive using Balena Etcher or Rufus
3. Boot from USB, select "Install Ubuntu Server"
4. Follow the installer — use the entire disk, enable OpenSSH
5. Create your user account and password
6. Reboot and remove the USB drive

After reboot, log in and update the system:

sudo apt update && sudo apt upgrade -y
sudo reboot

Step 2: Install NVIDIA Drivers + CUDA

The GPU driver is what lets your system talk to the hardware. CUDA is NVIDIA's parallel computing toolkit that AI frameworks depend on.

# Add NVIDIA package repository
sudo apt install -y linux-headers-$(uname -r)
sudo apt install -y nvidia-driver-550

# Reboot to load the driver
sudo reboot

# Verify the driver is working
nvidia-smi

You should see your GPU listed with driver version and memory information. If nvidia-smi shows your GPU, you're good.

# Install CUDA Toolkit
sudo apt install -y nvidia-cuda-toolkit

# Verify CUDA
nvcc --version

Step 3: Install Docker + NVIDIA Container Toolkit

Docker lets platforms run AI workloads in isolated containers. The NVIDIA Container Toolkit lets those containers access your GPU.

# Install Docker
curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker $USER
newgrp docker

# Install NVIDIA Container Toolkit
curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | \
  sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
  sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list

sudo apt update
sudo apt install -y nvidia-container-toolkit
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker

# Test GPU access from Docker
docker run --rm --gpus all nvidia/cuda:12.3.1-base-ubuntu22.04 nvidia-smi

If the last command shows your GPU inside a Docker container, everything is configured correctly.

Step 4: Choose and Register on a Platform

Now you need to connect your node to a decentralized network. Here's how to get started on the most beginner-friendly option:

Option A: Akash Network (Easiest)

1. Install the Akash provider software following their official docs
2. Create an Akash wallet and fund it with a small amount of AKT
3. Register your node as a provider
4. Set your pricing (start at market rate, adjust based on demand)
5. Your node will start receiving workloads automatically

Option B: io.net (Good earnings)

1. Create an account at io.net
2. Download and install the io.net worker software
3. Connect your GPU and complete the verification benchmark
4. Your node joins the GPU cluster and starts earning IO tokens

Option C: Bittensor (Highest potential, most complex)

1. Install the Bittensor CLI: pip3 install bittensor
2. Create a coldkey and hotkey wallet
3. Choose a subnet to mine on (each has different requirements)
4. Register your miner on the subnet
5. Run the subnet-specific mining software

Step 5: Monitor and Optimize

Your node is live. Here's what to do next:

• Monitor GPU usage: Run watch -n 1 nvidia-smi to see real-time GPU utilization
• Check earnings: Each platform has a dashboard showing your accumulated rewards
• Set up alerts: Use a simple cron job or monitoring tool to alert you if the node goes offline
• Keep software updated: Regularly update your OS, drivers, and platform software
• Track expenses: Log electricity usage and costs for ROI tracking and taxes

Common Issues

Problem	Solution
nvidia-smi not found	Driver not installed or Secure Boot blocking it. Reinstall driver with Secure Boot disabled.
Docker can't access GPU	NVIDIA Container Toolkit not configured. Re-run nvidia-ctk runtime configure and restart Docker.
Low utilization / no workloads	Normal during low demand. Ensure your pricing is competitive and uptime is 99%+.
High GPU temperature (85C+)	Improve case airflow, add fans, or reduce GPU power limit with nvidia-smi -pl 300.

What's Next?

Congratulations — you're now a decentralized AI compute provider. From here:

• Run your node for 30 days and track actual earnings vs. costs
• If profitable, consider adding a second GPU or building a dedicated multi-GPU rig
• Join your platform's community (Discord/Telegram) for tips and support
• Read our cost analysis to understand the long-term economics