domino logo
Latest (5.8)
  • About Domino
  • Architecture
  • Domino infrastructure
  • Kubernetes infrastructure
  • Installation
  • Configuration reference
  • Security and compliance
  • Authentication and authorization
  • Projects
  • Data Planes
  • Manage data
  • Models
  • Environments
  • Monitoring
  • Audits and logs
  • Control Center
  • Notifications
  • Manage costs
  • Disaster recovery
  • Backup and restore
  • Troubleshooting
  • Get help
  • Run the admin toolkit
  • MongoDB console
  • Send feedback
domino logo
About Domino
Domino Data LabKnowledge BaseData Science BlogTraining
>
Admin guide
>
Kubernetes infrastructure
>
NVIDIA DGX in Domino
>
NVIDIA DGX best practices

NVIDIA DGX best practices

  1. Build Node

    Domino recommends you do not use a DGX GPU as a build node for environments. Instead, opt for a CPU resource as part of your overall Domino architecture.

  2. Splitting GPUs per Tier

    Domino recommends providing several GPU tiers with different numbers of GPUs in each tier. For example, 1, 2, 4, and 8 GPU hardware tiers as different training jobs can take use of single or parallel GPU usage and consuming a whole DGX box for one workload might not be feasible in your environment.

  3. Governance

    After splitting up hardware tiers, access can be global or, alternatively, limited to specific organizations. Domino recommends that you ensure that the right organizations have GPU Hardware Tier access, or are restricted, to ensure availability for critical work, and/or to prevent the unauthorized use of GPU tiers.

Domino Data Lab
Knowledge Base
Data Science Blog
Training
Copyright © 2023 Domino Data Lab. All rights reserved.