CNCF [Cloud Native Computing Foundation] Courses

Building a Multi-Cluster Privately Hosted LLM Serving Platform on Kubernetes

Explore the architecture and challenges of building a cloud-agnostic LLM serving platform on Kubernetes, including control plane, dataplane, and serverless inference using KEDA for auto-scaling.

• YouTube
• 26 minutes
• On-Demand
• Free Video

How We Power the Largest AI Deployments on the Planet - Running Virtual Clusters at Scale

Explore challenges and solutions for managing large-scale Kubernetes clusters on bare metal, focusing on security, GPU provisioning, and scalability for AI workloads at CoreWeave and Loft Labs.

• YouTube
• 25 minutes
• On-Demand
• Free Video

Cluster Operations as a Service: Introducing LLM-Backed Controllers

Revolutionize cluster management with LLM-backed controllers. Deploy, scale, and secure Kubernetes clusters using natural language. Explore novel use cases, from chaos scenarios to security scanning and continuous improvement.

• YouTube
• 31 minutes
• On-Demand
• Free Video

Improving GPU Utilization and Accelerating Model Training with Kubernetes Scheduling Framework and NRI

Optimize GPU usage in K8s for AI workloads: implement GPU-sharing scheduling for efficient inference and topology-aware scheduling for accelerated distributed training using Scheduling Framework and NRI.

• YouTube
• 24 minutes
• On-Demand
• Free Video

Scheduling Jupyter Notebooks Using Airflow on Kubernetes

Learn to schedule Jupyter notebooks using Airflow on Kubernetes, enabling scalable data pipelines and machine learning workflows with enhanced security and resource management.

• YouTube
• 28 minutes
• On-Demand
• Free Video

Batch Systems in Production with Kueue - Multi-Tenancy and Fungibility

Explore Kueue, a cloud-native job scheduler for Kubernetes, enabling multi-tenant batch systems with efficient resource management, fair use, and support for various workloads and frameworks.

• YouTube
• 33 minutes
• On-Demand
• Free Video

Training Foundation Model Workloads on Kubernetes at Scale with MCAD

Explore how IBM's Vela supercomputer uses MCAD to queue custom resources for large-scale AI training on Kubernetes, enabling flexible frameworks and fault tolerance across hundreds of GPUs.

• YouTube
• 26 minutes
• On-Demand
• Free Video

Keep HPC Running - SRE's Guide to Supporting GPUs on Kubernetes

Learn best practices for SRE teams supporting GPU-enabled Kubernetes clusters for HPC and AI workloads, including key metrics, monitoring tools, and operational strategies.

• YouTube
• 22 minutes
• On-Demand
• Free Video

Training Large Language Models on Kubernetes

Explore challenges and best practices for training Large Language Models on Kubernetes, including optimization techniques, resource management, and benchmarks comparing K8s to bare metal environments.

• YouTube
• 28 minutes
• On-Demand
• Free Video

Constructing Kubeflow Super Nodes with Karpenter

Learn how Karpenter optimizes AI/ML platforms on Kubernetes by automatically selecting and launching ideal compute resources for Kubeflow Notebooks and Pipelines, enhancing efficiency and scalability.

• YouTube
• 29 minutes
• On-Demand
• Free Video

Poisoned Pickles - Security Risks and Protections for Serialized ML Models

Explore the security risks of using pickle for ML model serialization and learn strategies to protect against poisoned pickles in data science pipelines.

• YouTube
• 27 minutes
• On-Demand
• Free Video

Exploit Parallelism for AI Workloads with WASM and OpenMP

Explore optimizing AI workloads using WebAssembly and OpenMP for multithreaded, shared memory, and vectorized operations, focusing on challenges and benefits in cloud computing environments.

• YouTube
• 25 minutes
• On-Demand
• Free Video

Running Linux-Based Containers on Wasm and Browser with Container2wasm Converter

Learn to run Linux containers on WebAssembly and browsers using container2wasm converter. Explore its implementation, functionality, and integration with community tools.

• YouTube
• 28 minutes
• On-Demand
• Free Video

WebAssembly Component Model - Enhancing Security, Productivity, and Green Computing

Explore WebAssembly's component model for enhanced security, faster CVE resolution, and reduced environmental impact in application development.

• YouTube
• 31 minutes
• On-Demand
• Free Video

Navigating Strategies for Debugging WebAssembly

Explore effective strategies for debugging WebAssembly applications, including open-source tools and compile-time approaches, with insights into LLVM Intermediate Representation for improved development.

• YouTube
• 26 minutes
• On-Demand
• Free Video