DRA Driver for NVIDIA GPUs Overview

The DRA Driver for NVIDIA GPUs is an implementation of the Kubernetes Dynamic Resource Allocation (DRA) framework that exposes NVIDIA GPUs and Multi-Node NVLink as manageable resources.

Description

The DRA Driver for NVIDIA GPUs reflects the capabilities of underlying GPU hardware into standardized APIs that high-level Kubernetes components can use to unlock advanced features. It replaces the traditional, limited device plugin API with a richer, declarative interface.

The DRA Driver for NVIDIA GPUs enables controlled GPU sharing, dynamic MIG (Multi-Instance GPU) repurposing, and multi-node interconnect modeling (e.g., for GB200). It is designed specifically to meet the high-performance demands of modern AI/ML workloads at scale.

The container components are ready for commercial/non-commercial use.

License/Terms of Use

Use of this software is governed by the NVIDIA Software License Agreement and Product-Specific Terms for NVIDIA AI Products.

Reference(s):

The DRA Driver for NVIDIA GPUs community docs: https://github.com/kubernetes-sigs/dra-driver-nvidia-gpu/tree/main/docs

Security Common Vulnerabilities and Exposures (CVEs)

Please review the Security Scanning tab on NGC to view the latest security scan results. For certain open-source vulnerabilities listed in the scan results, NVIDIA provides a response in the form of a Vulnerability Exploitability eXchange (VEX) document. The VEX information can be reviewed and downloaded from the Security Scanning tab.

Get Help

Support for the upstream project is community-based. For documentation, issues, and project updates, refer to the GitHub repository and open issues.