Data Centers of the Future: Adapting Infrastructure for AI-ML Computing
Open Compute Project via YouTube
Most AI Pilots Fail to Scale. MIT Sloan Teaches You Why — and How to Fix It
Launch a New Career with Certificates from Google, IBM & Microsoft
Overview
Google, IBM & Meta Certificates — All 10,000+ Courses at 40% Off
One annual plan covers every course and certificate on Coursera. 40% off for a limited time.
Get Full Access
Explore a 13-minute conference talk from Google experts Varun Sakalkar and Mike Lau discussing how AI-ML technology evolution shapes the future of data center infrastructure. Dive into Google's computing infrastructure development, focusing on system-level architecture and the critical co-design approach from chip to data center scale. Learn about adapting infrastructure for high-power chips while maintaining warehouse-scale topology availability. Discover new developments in power delivery, liquid cooling, automation, serviceability, and data center architecture supporting next-generation AI-ML technologies. Gain insights into business and energy trends, onsite power solutions, and the concept of viewing data centers as machines, all through the lens of Google's ML mindset and innovative approaches to server and storage solutions.
Syllabus
Intro
Codesign
Scaling
Availability
High Power
Servers and Storage
ML Systems
Next Generation Power
ML Mindset
Data Center as Machine
Business Trends
Energy Trends
OnSite Power
Conclusion
Taught by
Open Compute Project