CNCF [Cloud Native Computing Foundation] Courses

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

How CoreWeave and Loft Labs use virtual Kubernetes clusters to deliver on-demand, serverless infrastructure for large AI workloads.

• YouTube
• 25 minutes
• Self-Paced
• Free Video

Cluster Operations as a Service: Introducing LLM-Backed Controllers

Explore Kubernetes controllers that use large language models to turn conversational requests into cluster operations, from workload deployment to recovery and security checks.

• YouTube
• 31 minutes
• Self-Paced
• Free Video

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

Learn how Kubernetes GPU sharing and topology-aware scheduling reduce inference resource waste and accelerate distributed model training.

• YouTube
• 24 minutes
• Self-Paced
• Free Video

Scheduling Jupyter Notebooks Using Airflow on Kubernetes

Learn how Kubernetes operators, autoscalers, and pod identities work with Airflow to schedule Jupyter notebooks and share execution results.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

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

See how Kueue uses quotas, borrowing, and preemption to share accelerator resources fairly across teams while keeping production batch workloads moving.

• YouTube
• 33 minutes
• Self-Paced
• Free Video

Training Foundation Model Workloads on Kubernetes at Scale with MCAD

See how MCAD queues custom resources and works with Kubernetes schedulers to support fault-tolerant foundation model training across hundreds of GPUs.

• YouTube
• 26 minutes
• Self-Paced
• Free Video

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

A practical SRE guide to monitoring GPU health in Kubernetes, using node problem detection and NVIDIA metrics exposed to Prometheus.

• YouTube
• 22 minutes
• Self-Paced
• Free Video

Training Large Language Models on Kubernetes

Learn how to pre-train large language models on Kubernetes, with guidance on cluster configuration and benchmarks against bare-metal training.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

Constructing Kubeflow Super Nodes with Karpenter

See how Karpenter selects machine architectures for Kubeflow Notebooks and Pipelines running machine learning jobs on Kubernetes.

• YouTube
• 29 minutes
• Self-Paced
• Free Video

Poisoned Pickles - Security Risks and Protections for Serialized ML Models

Learn how pickle-serialized machine learning models can be poisoned to inject code into ML pipelines, and how to protect against these attacks.

• YouTube
• 27 minutes
• Self-Paced
• Free Video

Exploit Parallelism for AI Workloads with WASM and OpenMP

A CNN kernel example shows how WASI threads and OpenMP can bring shared-memory parallelism and vectorization to WebAssembly AI workloads.

• YouTube
• 25 minutes
• Self-Paced
• Free Video

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

Learn how container2wasm converts Linux-based containers to run on WebAssembly runtimes and in browsers.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

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

See how WebAssembly components isolate dependencies so a vulnerable library can be rebuilt and relinked without rebuilding the rest of an application.

• YouTube
• 31 minutes
• Self-Paced
• Free Video

Navigating Strategies for Debugging WebAssembly

Explores how LLVM intermediate representation and compile-time analysis can help developers debug WebAssembly applications.

• YouTube
• 26 minutes
• Self-Paced
• Free Video

Orchestrating WebAssembly - Reconciliation Loops Beyond Kubernetes

A technical look at how wasmCloud’s Wadm uses reconciliation loops to orchestrate WebAssembly applications across regions and heterogeneous compute environments.

• YouTube
• 29 minutes
• Self-Paced
• Free Video