MIG on DGX Station

Basic idea

Multi-Instance GPU (MIG) lets you partition a single NVIDIA B300 GPU on your DGX Station (GB300 Ultra) into multiple smaller GPU instances. Each instance has dedicated memory and compute, so you can run multiple workloads or users on one physical GPU without sharing memory. This playbook walks you through enabling MIG, creating a B300 MIG layout, and using the instances from bare-metal apps or containers.

MIG is controlled via nvidia-smi: you enable MIG mode, then create GPU and compute instances using B300 profile IDs (e.g. 1g.34gb, 2g.67gb, 7g.269gb). When you no longer need partitioning, you disable MIG to restore full-GPU and NVLink P2P.

What you'll accomplish

You will have MIG enabled and configured on your DGX Station B300 GPUs and know how to use the instances.

• Enable MIG on all B300 GPUs or on a per-GPU basis.
• Create a MIG layout using B300 profile IDs (with a known-good example for multiple GPUs).
• Verify the layout with nvidia-smi -L and sudo nvidia-smi mig -lgi / -lci.
• Run workloads by setting CUDA_VISIBLE_DEVICES to a MIG UUID or by using the container/Kubernetes flows from the MIG User Guide.
• Disable MIG when you need full-GPU mode and NVLink again.

What to know before starting

• Basic Linux command line and use of sudo.
• Familiarity with nvidia-smi and GPU indices.
• Optional: understanding of CUDA_VISIBLE_DEVICES and containers if you plan to run workloads on MIG instances.

Prerequisites

Hardware:

• NVIDIA DGX Station with GB300 Ultra Superchip (B300 GPUs).
• No additional storage requirement for MIG configuration itself.

Software:

• NVIDIA driver and nvidia-smi installed and working: nvidia-smi. Use a driver version that supports MIG on B300 (see Troubleshooting for version guidance; if nvidia-smi -mig 1 reports "MIG mode not supported" or similar, the driver may be too old).
• Root or sudo access to run nvidia-smi -mig 1, -mig 0, and nvidia-smi mig -cgi ... -C
• For containers/K8s: nvidia-container-toolkit and MIG support as described in the MIG User Guide

Ancillary files

This playbook does not use repository assets; all steps use nvidia-smi and MIG commands on the DGX Station. For container and Kubernetes setup, use the official MIG User Guide (Getting Started with MIG and Kubernetes sections).

Time & risk

• Estimated time: About 15 minutes to enable MIG, create a layout, and verify. Layout design (which profiles per GPU) may take longer if you customize.
• Risk level: Low to Medium
  • Enabling or disabling MIG requires sudo and affects all workloads on that GPU.
  • Disabling MIG removes all MIG instances; ensure Fabric Manager is running on DGX/HGX B200/B300 so NVLink/NVSwitch re-initialize correctly.
• Rollback: Destroy all MIG instances with sudo nvidia-smi mig -dci -i N and sudo nvidia-smi mig -dgi -i N for each GPU index N, then run sudo nvidia-smi -mig 0 to disable MIG and return to a single full-GPU instance per GB300. Ensure Fabric Manager is running after disabling MIG: sudo systemctl status nvidia-fabricmanager (start if needed: sudo systemctl start nvidia-fabricmanager).
• Last Updated: 03/02/2026
  • First publication.