Running Large-Scale Machine Learning on Google Kubernetes Engine

Discover how to run large-scale machine learning on Google Kubernetes Engine (GKE) in this 29-minute conference talk from Google Cloud Next. Learn about the challenges of training and serving large language models (LLMs) at scale, and explore how companies leverage GKE to build extensive ML models. Gain insights into GKE's solutions for common challenges in training large AI models, and uncover best practices for training and serving large-scale ML models on the platform. Speakers Nathan Beach and Alex Zakonov discuss GKE investments, ML training enablement, A3 VM, job queuing, GKE infrastructure, and key features that make GKE an ideal choice for large-scale machine learning operations.

Introduction
Challenges
Why customers use GKE
GKE Investments
How GKE enables ML training
A3 VM
Job queuing
GKE infrastructure
GKE features