Manually Install, Update, and Uninstall on Ubuntu (Remote)#

Learn how to manually install, update, or remove NVIDIA AI Workbench on a remote Ubuntu system.

Agree to the NVIDIA AI Product Agreement

Before downloading, please read the NVIDIA AI Product Agreement, the NVIDIA AI Workbench Shared Security Model, and our Data Privacy Policy.

By downloading, installing, or using the NVIDIA AI Workbench software, you agree to the terms of the NVIDIA AI Product Agreement (EULA). If you do not agree to the terms of the EULA, you are not authorized to download, install, or use NVIDIA AI Workbench. Before downloading and using NVIDIA AI Workbench, please read our Shared Security Model.

Prerequisites for Your Remote Ubuntu System#

  • A remote Ubuntu 24.04 or 22.04 system

    • SSH access with a public/private key pair as a non-root user with sudo privileges

  • System resources:

    • 500 MB of disk space for the AI Workbench application

    • 30-40 GB of disk space for containers

    • 16 GB of RAM

Install AI Workbench on Remote Ubuntu#

You must use the CLI to install AI Workbench on a remote system. You can use the CLI interactively or non-interactively.

Interactive Installation#

  1. SSH into the Remote as a non-root user with sudo privileges

    ssh -i <local-path-to-private-key> <non-root-username>@<remote-host>

  2. Copy, paste and run the installation command in the remote terminal

    1mkdir -p $HOME/.nvwb/bin && \
2curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-$(uname)-$(uname -m) --output $HOME/.nvwb/bin/nvwb-cli && \
3chmod +x $HOME/.nvwb/bin/nvwb-cli && \
4sudo -E $HOME/.nvwb/bin/nvwb-cli install

  3. Use the Arrow Keys and Press Enter to Follow the Installation Prompts in the Terminal

    • If prompted, choose between Docker or Podman as the container runtime. ( see AI Workbench Container Runtimes)

    • If prompted, install the NVIDIA GPU drivers

  4. Connect to the Remote System in AI Workbench

Non-Interactive Installation#

Use non-interactive installation if you’re automating setup or scripting deployment.

  1. SSH into the Remote as a non-root user with sudo privileges

    ssh -i <local-path-to-private-key> <non-root-username>@<remote-host>

  2. Download and install the AI Workbench CLI

    1mkdir -p $HOME/.nvwb/bin && \
2curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-$(uname)-$(uname -m) --output $HOME/.nvwb/bin/nvwb-cli && \
3chmod +x $HOME/.nvwb/bin/nvwb-cli

  3. Run the non-interactive installer with all required flags

    1sudo $HOME/.nvwb/bin/nvwb-cli install \
2  --noninteractive \
3  --accept \
4  --podman \
5  --drivers \
6  --uid 1000 \
7  --gid 1000

    Replace --podman with --docker if you prefer Docker as the container runtime.

    Note

    • You must specify exactly one of --podman or --docker.

    • You must include --accept to agree to the NVIDIA AI Product Agreement.

    • --drivers is optional and only applies to Linux systems with NVIDIA GPUs.

    • --uid and --gid are required if installing as root for another user.

  4. Restart your shell session

    The nvwb CLI will be available after your shell session is restarted.

Update AI Workbench on a Remote System#

Updating AI Workbench on a remote system is basically the same as installing it. You just pull the latest binaries and run the install command again.

  1. SSH into the Remote as the same non-root user that did the installation

  2. Download and Install the Latest AI Workbench Version with the following command:

    1curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-$(uname)-$(uname -m) --output $HOME/.nvwb/bin/nvwb-cli && \
2chmod +x $HOME/.nvwb/bin/nvwb-cli && \
3sudo -E $HOME/.nvwb/bin/nvwb-cli install

  3. Follow the prompts in the text-based interface to finish the update

Uninstall AI Workbench on a Remote System#

  1. SSH into the Remote as the same non-root user that did the installation.

  2. Run the Uninstall Command, Type y and Press Enter to Confirm

    sudo -E $HOME/.nvwb/bin/nvwb-cli uninstall

  3. (Optional) Remove All AI Workbench Related Files and Data

    • Remove AI Workbench configuration files:

      rm -rf $HOME/.nvwb

    • Remove all project repositories (if no longer needed):

      rm -rf $HOME/nvidia-workbench

FAQs#

What Gets Installed/Updated and Where?#

  • The application binaries: nvwb-cli, wb-svc, and credential-manager into $HOME/.nvwb/bin

  • Git: Installed or updated system wide

  • Docker or Podman: Installed or updated system wide

  • (If GPU): NVIDIA GPU Drivers, if not already installed

  • (If GPU): NVIDIA Container Toolkit, installed or updated system wide

Do I install the Desktop App on a remote?#

No. The Desktop App is not supported on a remote system.

  • You cannot install on a remote with either the Debian package or the AppImage.

  • You must use the CLI for the remote install.

How do I do a non-interactive install using the CLI?#

Next Steps#