Skip to navigation

NVIDIA Run:ai Self-Hosted Product Documentation

View as Markdown

NVIDIA Run:ai is an open-source AI workload and GPU orchestration platform for managing shared AI infrastructure on Kubernetes.

NVIDIA Run:ai provides a consistent way to run development, training, and inference workloads across Kubernetes environments, whether deployed on-premises, in the cloud, or across hybrid infrastructure. It combines GPU scheduling, resource management, workload orchestration, governance, and observability to help organizations share GPU resources efficiently across teams and clusters.

Support-Entitled Release

NVIDIA Run:ai is available as open source. The NVIDIA Run:ai Support-Entitled Release is available for organizations that require NVIDIA enterprise support and an SLA.

Getting Started

Get started with NVIDIA Run:ai

Follow the quick start guides to deploy infrastructure, configure the platform, and run your first workloads.

Install NVIDIA Run:ai

Review system requirements and deploy NVIDIA Run:ai using the guided installer or Helm.

Local Development

Build NVIDIA Run:ai and run the full stack locally for development.

Using NVIDIA Run:ai

Manage shared AI infrastructure

Organize GPU resources across departments and projects, allocate resources to teams, and control how shared capacity is used.

Optimize GPU resources

Share GPU capacity across workloads and use scheduling, priorities, and resource allocation controls to improve utilization.

Run AI workloads

Run development, training, distributed training, and inference workloads on shared Kubernetes infrastructure.

Operate NVIDIA Run:ai at scale

Manage multiple clusters, configure platform services, monitor system health and resource utilization, and maintain your NVIDIA Run:ai deployment.

Resources

Videos

Watch NVIDIA Run:ai platform overviews, walkthroughs, and task-based demonstrations.

Blogs

Explore NVIDIA Run:ai articles and technical content.

NVIDIA On-Demand Sessions

Watch NVIDIA Run:ai sessions, presentations, and technical talks.

What's New

See the features and enhancements in the latest release.

Upgrade

Upgrade an existing NVIDIA Run:ai deployment to a new version.

Product Version Life Cycle

Plan upgrades based on support timelines.