In this course, you'll learn to deploy and manage NVIDIA AI Enterprise on Bare-Metal Kubernetes environments. The training follows a scenario-based approach, starting with Kubernetes deployment on bare-metal infrastructure, followed by NVIDIA AI Enterprise implementation, comprehensive GPU validation to ensure readiness for accelerated AI workloads, and concluding with a deployment of a standard Retrieval Augmented Generation (RAG) use case.

• NVIDIA AI Enterprise on Bare-Metal Overview: This section introduces NVIDIA AI Enterprise and its implementation on bare-metal infrastructure. You'll learn about NVIDIA AI Blueprints, and NVIDIA Inference Microservice (NIM) for LLMs. • NVIDIA AI Enterprise Bare-Metal Deployment Overview: This section covers deployment methods, and hardware and software prerequisites for bare-metal implementations. We walk you through preparing a bare-metal platform for NVIDIA AI Enterprise deployment. • platform for NVIDIA AI Enterprise deployment. • NVIDIA AI Enterprise Deployment: This section explores NVIDIA AI Enterprise platform prerequisites and NGC resources including Catalog, CLI, and API. We walk you through deploying NVIDIA AI Enterprise on Kubernetes in a bare-metal environment. • Management & Monitoring: This section addresses tools and techniques for monitoring, scaling, and maintaining NVIDIA AI Enterprise deployments. You'll learn about workload management options including Kubernetes and Base Command Manager. • Reference Use Case: This section demonstrates an image recognition and practical enterprise RAG workflows deployment. • Troubleshooting and Support: This section covers common issues, resolution techniques, and support resources for bare-metal deployments. You'll learn troubleshooting methodologies specifically tailored for bare-metal NVIDIA AI Enterprise implementations.

• Recall the key features and benefits of deploying NVIDIA AI Enterprise on bare-metal. • Identify the prerequisites for deploying NVIDIA AI Enterprise on bare-metal. • Implement the steps to deploy NVIDIA AI Enterprise on bare-metal with K8s. • Construct a basic K8s cluster on bare-metal and configure it for use with NVIDIA AI Enterprise. • Apply NVIDIA AI Enterprise deployment techniques on K8s using Helm charts. • Utilize the NVIDIA GPU Operator for GPU configuration deployment. • Evaluate GPU deployment effectiveness by implementing and testing an example training job. • Analyze an enterprise RAG use case through practical workflow deployment. • Monitor and manage NVIDIA AI Enterprise deployments on K8s. • Design scaling strategies and update procedures for NVIDIA AI Enterprise deployments on K8s. • Create solutions for troubleshooting common issues and effectively accessing support for NVIDIA AI Enterprise on bare-metal K8s deployments.

• IT professionals • DevOps/MLOps engineers • Anyone responsible for deploying and managing AI infrastructure on bare-metal.

To gain the most value from this course, the target audience should have knowledge of the following domains: • Working knowledge of data center infrastructure: - Servers - Storage - Networking - GPUs - Operating systems • Familiarity with containerization and Kubernetes: - Docker - Kubernetes - Cloud Native Computing Foundation (CNCF)