NGC CLI installation: Worked Example — Control Plane Installation and Configuration (NVIDIA-Certified Professional: AI Infrastructure)

NGC CLI Installation Worked Example for NVIDIA-Certified Professional: AI Infrastructure The NVIDIA GPU Cloud (NGC) Command Line Interface (CLI) is a...

NGC CLI Installation Worked Example for NVIDIA-Certified Professional: AI Infrastructure

The NVIDIA GPU Cloud (NGC) Command Line Interface (CLI) is a critical tool for managing NVIDIA AI software resources, including containers, models, and workflows. Proper installation and configuration of the NGC CLI is essential for AI infrastructure professionals to efficiently deploy and maintain AI workloads.

Scenario Overview

Assume you are an AI infrastructure engineer tasked with installing the NGC CLI on a Linux-based AI compute node. The goal is to enable seamless access to NVIDIA NGC resources, ensuring the CLI is configured correctly for authentication and integration with GPU-enabled containers.

Step 1: Prepare the System Environment

Ensure the system has Python 3.6+ installed, as the NGC CLI requires Python.
Verify network connectivity to ngc.nvidia.com for authentication and resource access.
Confirm that NVIDIA GPU drivers and the NVIDIA Container Toolkit are installed and functional.

Step 2: Download the NGC CLI

Obtain the latest NGC CLI binary from the official NVIDIA source:

Navigate to the NGC setup page to generate an API key if you do not have one.
Download the CLI binary using wget or curl:

Example Command

wget https://ngc.nvidia.com/downloads/ngccli_linux.zip

Step 3: Install the NGC CLI

Unzip the downloaded archive:

unzip ngccli_linux.zip

Move the extracted ngc binary to a directory in your PATH, such as /usr/local/bin:

sudo mv ngc /usr/local/bin/

Make the binary executable:

sudo chmod +x /usr/local/bin/ngc

Step 4: Authenticate the NGC CLI

Authentication is required to access private and public NGC resources.

Use the API key generated from the NGC website:

ngc config set

This command will prompt you to enter your API key and optionally set a default output format and region.

Step 5: Verify Installation and Configuration

Check the NGC CLI version:

ngc --version

List available NGC resources to confirm connectivity:

ngc registry image list

Run a test pull of a GPU-enabled container image to verify GPU access through Docker:

docker run --gpus all nvcr.io/nvidia/cuda:11.8.0-base-ubuntu20.04 nvidia-smi

This command should display GPU information, confirming that the NGC CLI and NVIDIA container toolkit are working together.

Step 6: Troubleshooting Tips

If authentication fails, verify the API key and network access.
Ensure Docker is configured with the NVIDIA runtime by checking /etc/docker/daemon.json.
Confirm GPU drivers and CUDA toolkit versions are compatible with the container images.

Summary

This worked example demonstrated the step-by-step installation and configuration of the NGC CLI on a Linux AI compute node. Mastery of this process is vital for NVIDIA-Certified Professionals in AI Infrastructure to efficiently manage AI workloads and leverage NVIDIA's AI software ecosystem.

More in this topic

Ready to test your knowledge?

Put what you've learned into practice with a quick quiz and track your progress.

Test your knowledge →