Overview

Optimize AI & Data Science Workloads (Red Hat OpenShift) (Latest)

Welcome to NVIDIA LaunchPad and the NVIDIA AI Enterprise Evaluation of RedHat OpenShift! Within your journey here, you will access leading AI software and infrastructure, enabling your Enterprise to speed up the development and deployment of modern, data-driven applications. We’ll dive into the LaunchPad infrastructure and take you through the IT Administrator user persona journey, which accelerates supporting AI initiatives within an enterprise.

First, let’s start with the NVIDIA-Certified System ™ LaunchPad infrastructure. NVIDIA AI Enterprise compatible servers power your LaunchPad journey; these are NVIDIA-Certified and support Red Hat OpenShift 4.9 or later. NVIDIA-Certified Systems brings together NVIDIA GPUs and NVIDIA networking in servers from leading NVIDIA partners in optimized configurations. These servers are validated for performance, manageability, security, and scalability and backed by NVIDIA and partners’ enterprise-grade support.

NVIDIA AI Enterprise is an end-to-end, cloud-native suite of AI and data science applications and frameworks optimized and exclusively certified by NVIDIA to run Red Hat OpenShift with NVIDIA-Certified Systems. It includes key enabling technologies and software from NVIDIA for rapid deployment, management, and scaling of AI workloads in the modern hybrid cloud.

openshift-it-001.jpg

Within this IT Administrator LaunchPad journey, you will walk through the steps to install, configure, and validate the NVIDIA GPU Operator on Red Hat OpenShift Container Platform (OCP) running on bare metal.

As part of this lab, you will perform the following tasks:

  • Install the Node Feature Discovery (NFD) Operator

  • Create NVIDIA NGC pull secret

  • Install the NVIDIA GPU Operator

  • Create a GPU Cluster Policy Instance

  • Build Grafana dashboards to monitor GPU usage

  • Run Sample GPU workloads

Important

To assist you in our LaunchPad journey, there are a couple of important links on the left-hand navigation pane of this page. In the next step, you will use the top LaunchPad OCP link.

Note

Unless otherwise noted, all steps will be performed in the OpenShift Web Console. Alternatively, if you would like to execute CLI commands a System Console located in the left-hand navigation pane. The kubeconfig file connecting to the OpenShift cluster is stored in the nvidia user’s home directory.

© Copyright 2022-2023, NVIDIA. Last updated on Jan 10, 2023.