DGX Resources
NVIDIA Docs Hub  DGX Resources

Welcome

Thanks for purchasing NVIDIA DGX Products. Here you will find resources that will help you get started.

Getting started

Follow the instructions that came in your entitlement email to register for an NVIDIA Enterprise account. After successful registration and account activation, you’ll get login access to the following NVIDIA software portals.

  • NVIDIA NGC: Provides access to all enterprise software, services, and management tools included in your purchase. 
  • NVIDIA Enterprise Support Portal: Provides access to support services for the products you’ve purchased. 
  • NVIDIA Licensing Portal: Provides access to your entitlements and options for managing your license keys and license servers. 

Your purchase may include the following DGX software stack

  • NVIDIA Mission Control (NMC), includes 
    • Base Command Manager (BCM) 
      • Powerful and flexible AI infrastructure provisioning and management software for HPC and AI workloads deployed on-prem, in the cloud or at the edge. 
    • Run:ai  
      • AI workload scheduler with a tenant control plane on-prem or in the cloud. Supports GPU fractioning to fine-tune workload deployments, and heterogeneous AI cluster control with quotas, priorities and policies. 
    • NVIDIA Unified Fabric Manage (UFM)  
      • Helps manage InfiniBand scale-out computing environments and may require additional hardware appliances.
  • NVIDIA DGX OS  
    • Provides a customized installation of Ubuntu Linux with system-specific optimizations and configurations, additional drivers, and diagnostic and monitoring tools. It provides a stable, fully tested, and supported OS to run AI, machine learning, and analytics applications on DGX Supercomputers. The software packages included in the DGX OS software stack are also available and installable on top of a vanilla Ubuntu distribution and Red Hat Enterprise Linux. Refer to NVIDIA Base OS for more details.
       
  • Magnum IO 
    • Utilizes storage IO, network IO, in-network compute, and IO management to simplify and speed up data movement, access, and management for multi-GPU, multi-node systems. Learn more about it here. 
  • NVIDIA AI Enterprise 
    • A suite of software optimized to streamline AI development and deployment. Learn more about it here. 

To download, install and activate the software that comes with your purchase, follow these steps:

______________________________________________________________________________________________________

  • NVIDIA Mission Control (NMC)
    • Base Command Manager: 
      • Using the PAK ID on the entitlement email, generate your Base Command Manager Product Key from the  NVIDIA Licensing Portal (NLP).  Within the NLP, click the “Base Command Manager” section on the left navigation pane. Then on the Base Command Manager line item, click the 3 vertical dots under the ‘Actions’ column to expose a “Generate key” button. This key will be used to activate your license.
      • Download your ISO for Base Command Manager software by visiting the Base Command Manager Download site.  You will need to fill in the following details 
        • Product key 
        • Linux version 
        • Base Command Manager version (if not the most current version) 
      • Should you need support when installing your software, please visit the NVIDIA Enterprise Support Portal. 
        Once you’ve installed the software on your head node, use the 'request-license' command on the head node to activate the license for your cluster. Detailed instructions can be found in chapter 4 of the Installation Manual. 
    • Run:ai:
    • NVIDIA Unified Fabric Manager (UFM):  
      • Generating License File: 
        • Prepare a list of servers with the MAC address of each server on which you plan to install the UFM software. 
        • Go to NVIDIA’s NVIDIA Licensing Portal (NLP) and log in using your credentials. 
        • Click on the Network Entitlements tab. You'll see a list with the serial licenses of all your software products and software product license information and status. 
        • Select the license you want to activate and click on the “Actions” button. 
        • In the MAC Address field, enter the MAC address of the delegated license-registered host. If applicable, in the HA MAC Address field, enter your High Availability (HA) server MAC address. If you have more than one NIC installed on a UFM Server, use any of the MAC addresses. 
        • Click on Generate License File to create the license key file for the software. 
        • Click on Download License File and save it on your local computer.
        • If you replace your NIC or UFM server, repeat the process of generating the license to set new MAC addresses. You can only regenerate a license two times. To regenerate the license after that, contact NVIDIA Sales Administration at enterprisesupport@nvidia.com. 
      • Downloading UFM Software: 
        • Go to NVIDIA’s NVIDIA Licensing Portal (NLP) and log in using your credentials. 
        • Click on Software Downloads, filter the product family to UFM, and select the relevant version of the software. Click on Download. 
        • Save the file on your local drive. 
        • Click Close. 

____________________________________________________________________________________________________

  • DGX OS 
    • DGX OS is included in BCM 10. Just select the correct image for your system. 
    • If you want to re-image your systems, you can download the DGX OS ISO from the NVIDIA Licensing Portal following these steps: 
      • Go to the Download Center. 
      • Click [Server/Workstation] -> [DGX] and select All Downloads for your system. 
      • Click on the download link for the latest ISO release to go to the announcement. 
      • Download the ISO image that is referenced in the announcement and save it to your local disk. 
    • For Red Hat Enterprise Linux or vanilla Ubuntu deployments, including other cluster management systems, refer to the User Guides available from NVIDIA Base OS User Guides 
    • Keeping your DGX OS software and firmware up to date is the most important task for protecting your DGX systems. Security-related updates are available from the Ubuntu and NVIDIA repositories. Please refer to the User Guides for upgrade instructions. 

______________________________________________________________________________________________________

  • DGX System Software and Firmware 
    • NVIDIA DGX systems have several firmware components. It is important to keep these up to date to avoid security, software, or hardware issues. Please refer to the system firmware update guides listed in the NVIDIA DGX Systems Guides  for your system. 

_______________________________________________________________________________________________________

Stay in touch with our latest product updates and announcements

Connect and learn from DGX community.

There are various ways in which a DGX user can connect with the experts and the community

  • DGX User Forum: Join the DGX User Forum.  
    • Connect with other DGX users and the NVIDIA DGX Product team. Gain insights and tips to help you get the most out of your NVIDIA DGX System.  
  • DGX Information Sessions:
    • As a valued DGX customer, we are giving you direct access to our best practices and AI expertise through a series of live DGX information sessions. Get answers to your questions, with topics ranging from planning to deployment to on-going optimization, led by our own DGXperts - AI-fluent professionals who have deployed thousands of DGX systems like yours. If you registered for enterprise support, you will automatically be receiving these invitations. 
  • DGX ORIENTATION 
    • DGX Orientation is included with all DGX purchases. NVIDIA Professional Services (NVPS) offers a walkthrough of capabilities, software, and tools provided through the NVIDIA end-to-end accelerated stack, and includes best practices for getting the most of the DGX platform as quickly as possible. DGX Orientation is divided into two 4-hour on-demand sessions. Register to view the first on-demand session, and sign up for one of 4 monthly Q&A sessions to get live advice from the experts.