Use this documentation to install, update, or uninstall NVIDIA AI Workbench on your remote computer running Ubuntu 22.04. To review important considerations before you start, see Install AI Workbench.
Use this documentation to perform the following tasks:
Ubuntu 22.04
SSH access as a non-root user with
sudo
privileges by using a public/private key pair. The private key can’t be password-protected.500 MB of disk space for the AI Workbench application
30-40 GB of disk space for containers
16 GB of RAM
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.
Use the following procedure to install AI Workbench on your remote computer running Ubuntu.
From a terminal on your local computer, SSH to the remote computer. You can use a password or private key for this step.
WarningDo not ssh to the remote computer as the
root
user.Run the following command.
mkdir -p $HOME/.nvwb/bin && \ curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-$(uname)-$(uname -m) --output $HOME/.nvwb/bin/nvwb-cli && \ chmod +x $HOME/.nvwb/bin/nvwb-cli && \ sudo -E $HOME/.nvwb/bin/nvwb-cli install
A text-based interface starts.
The text-based interface guides you through the remote installation.
You can advance through the installation by using the arrow keys to select different options and pressing
Enter
to select options.If the computer has an NVIDIA GPU but no drivers installed, AI Workbench detects this and asks you if you want the drivers installed for you. If you choose yes, the drivers are installed and you reboot the computer in a later step.
Choose Docker or Podman as your container runtime. For more information, see Container runtime options. AI Workbench installs the option you select.
If you chose to have drivers installed during the installation, reboot the computer.
Now you can connect to the remote computer from your local AI Workbench.
After you install AI Workbench on a remote computer, you only have to connect to the remote computer from your local computer one time. After that, the remote computer appears as a Location whenever you open AI Workbench locally. To connect to the Remote Location, do the following steps:
Generate a passwordless ssh key and add the public key to the remote instance’s
.ssh/authorized_keys
file. Open a command terminal on your local computer and run the following code that corresponds to your operating system.NoteMake these changes first:
Change
local-user
to the username on your local computer.Change
remote-user
to the username on your remote computer.Change
10.123.4.56
to the IP address of your remote computer.
ssh-keygen -f "C:\Users\local-user\.ssh\id_rsa" -t rsa -N '""' type$env:USERPROFILE\.ssh\id_rsa.pub | ssh remote-user@10.123.4.56 "cat >> .ssh/authorized_keys"
ssh-keygen -f "/Users/local-user/.ssh/id_rsa" -t rsa -N "" ssh-copy-id remote-user@10.123.4.56
ssh-keygen -f "/home/local-user/.ssh/id_rsa" -t rsa -N "" ssh-copy-id remote-user@10.123.4.56
Open the AI Workbench desktop application on your local computer. Click Add Remote Location, and then enter the following information for your Remote Location:
Location Name
Description
Hostname or IP Address
SSH Port
SSH Username — The username on the remote computer.
SSH Key File — The private key file on your local computer.
Workbench Directory — The
.nvwb
directory on the remote computer. If you didn’t change the path, accept the default.
Click Add Location. The remote computer appears as a location in AI Workbench.
Use the following procedure to update AI Workbench on your remote computer running Ubuntu.
From a terminal on your local computer, SSH to the remote computer. You can use a password or private key for this step.
WarningDo not ssh to the remote computer as the
root
user.Run the following command to download the latest version of AI Workbench.
mkdir -p $HOME/.nvwb/bin && \ curl -L https://workbench.download.nvidia.com/stable/workbench-cli/$(curl -L -s https://workbench.download.nvidia.com/stable/workbench-cli/LATEST)/nvwb-cli-$(uname)-$(uname -m) --output $HOME/.nvwb/bin/nvwb-cli && \ chmod +x $HOME/.nvwb/bin/nvwb-cli && \ sudo -E $HOME/.nvwb/bin/nvwb-cli install
A text-based interface starts.
The text-based interface guides you through the update.
You can advance through the update by using the arrow keys to select different options and pressing
Enter
to select options.
Confirm Docker or Podman as your container runtime.
Now you can connect to the remote computer from your local AI Workbench.
Use the following procedure to uninstall AI Workbench from your remote computer running Ubuntu.
Drivers are not uninstalled, even if AI Workbench installed them for you.
From a terminal on your local computer, SSH to the remote computer. You can use a password or private key for this step.
WarningDo not ssh to the remote computer as the
root
user.Run the the following command.
sudo $HOME/.nvwb/bin/nvwb-cli uninstall
You are prompted to confirm that you want to uninstall AI Workbench. Type y and press enter.
Wait for the uninstall to complete before you proceed with the following steps.
(Optional) If you want to remove the application files, run the following command to remove the .nvwb folder.
rm -r $HOME/.nvwb
(Optional) Iif you want to remove your project repositories, run the following command.
rm -r $HOME/nvidia-workbench/
Do a Quickstart
Explore the NVIDIA Developer Forum
Learn the Concepts