Kubernetes Podcast from Google Kubernetes AI Conformance, with Janet Kuo
13 snips
Dec 17, 2025 Janet Kuo, a Staff Software Engineer at Google Cloud and key figure behind the Kubernetes AI Conformance program, dives into the new AI conformance initiative. She highlights how AI workloads differ from standard Kubernetes conformance, emphasizing their unique demands for stateful operations and hardware. Janet explains the introduction of the Dynamic Resource Allocation (DRA) API, designed to improve accelerator management for complex AI needs. She also shares insights on future standardizations and the importance of consistent platform capabilities.
AI Snips
Chapters
Transcript
Episode notes
GKE Scaled To 130,000 Nodes
- Google Cloud demonstrated a 130,000-node GKE cluster by re-architecting the control plane and replacing etcd with Spanner.
- This showcases Kubernetes evolution to meet massive AI and data workload scale.
AI Conformance Extends Kubernetes Guarantees
- Kubernetes AI conformance builds on regular conformance and requires Kubernetes conformance first.
- It extends guarantees to AI workload needs so platforms offer consistent AI behavior everywhere.
AI Workloads Demand Platform Guarantees
- AI workloads have distinct needs like specialized networking, accelerators, and stateful execution.
- The AI conformance program targets those platform-level capabilities rather than prescribing how to run workloads.