Run:ai shared responsibility (DGX Cloud)

NVIDIA manages:

Infrastructure health monitoring and availability
Kubernetes cluster deployment and updates
Software patches and vulnerability management
Security of cloud services and workload isolation

Customer manages:

User access and role assignments
Resource quota allocations per project
User workloads (create, modify, delete)
Data protection and backup
Network ingress CIDR range (changes require a TAM)
Reporting issues

Access:

Ingress to control plane via SSO — Port 443
Kubernetes control plane via customer-specified CIDR — Port 443
CIDR changes require engaging your NVIDIA Technical Account Manager (TAM)

Run:ai organization and roles Next: Run:ai CLI

Study tools

Mark this unit complete when you finish it.

Completed