Cluster API To Streamline Kubernetes Clusters Lifecycle Management at NVIDIA

Learn how NVIDIA consolidated its Kubernetes infrastructure management using ClusterAPI and bare-metal machines in this 12-minute conference talk from KubeCon + CloudNativeCon. Discover NVIDIA's implementation approach that incorporates heterogeneous machine types including CPU, GPU, TEGRA, and ARM architectures to enable seamless high-performance workloads. Explore how the team developed a custom ClusterAPI infrastructure provider to make Kubernetes cluster lifecycle management more efficient, enhanced ClusterAPI by wrapping templates into Helm charts for standardized deployments, and implemented GitOps workflows to ensure declarative, version-controlled cluster creation and management. Understand how this solution eliminated silos and inefficiencies that typically arise when teams independently provision and manage Kubernetes clusters, while reducing operational overhead and improving consistency across multi-environment deployments.