Compute Setup

x86 Platforms

  1. Prepare a NVIDIA-powered platform with the following minimum specs:

    • Ubuntu 20.04+

      • Experimental: WSL2 on Windows 11

    • 16 GB general RAM

    • Discrete NVIDIA GPU with the following specs:

      • supports CUDA 11.8 with Volta or newer

      • minimum 8GB of VRAM and recommended 12GB+

  2. Install Docker from official instructions (here)

  3. Install the latest NVIDIA GPU Driver from official instructions (here)

Jetson Platforms

  1. Install Jetpack including the nvidia-container package.

  2. After boot, confirm that you have installed Jetpack 5.1.2 by running the following command. Confirm that the output has the terms R35 (release), REVISION: 4.1.

    cat /etc/nv_tegra_release
  3. Run the following command to set the GPU and CPU clock to max. See Maximizing Jetson Performance for more details.

    sudo /usr/bin/jetson_clocks
  4. Run the following command to set the to power to MAX settings. See Power Mode Controls for more details.

    sudo /usr/sbin/nvpmodel -m 0
  5. Add your user to the docker group.

    sudo usermod -aG docker $USER
    newgrp docker
  6. Setup Docker.

    From the official Docker install instructions (here), install the docker-buildx-plugin.

    # Add Docker's official GPG key:
    sudo apt-get update
    sudo apt-get install ca-certificates curl gnupg
    sudo install -m 0755 -d /etc/apt/keyrings
    curl -fsSL | sudo gpg --dearmor -o /etc/apt/keyrings/docker.gpg
    sudo chmod a+r /etc/apt/keyrings/docker.gpg
    # Add the repository to Apt sources:
    echo \
    "deb [arch="$(dpkg --print-architecture)" signed-by=/etc/apt/keyrings/docker.gpg] \
    "$(. /etc/os-release && echo "$VERSION_CODENAME")" stable" | \
    sudo tee /etc/apt/sources.list.d/docker.list > /dev/null
    sudo apt-get update
    sudo apt install docker-buildx-plugin

Consider adding more storage to your Jetson for a better experience: Developer Environment Setup for Jetson