Watch a 14-minute keynote presentation from NVIDIA's Ian Buck at the Open Compute Project where he unveils the GB200 NVL72's design principles and architecture, specifically engineered for trillion parameter AI workloads. Explore how this new GPU architecture and rack-scale design addresses critical challenges in power consumption, cooling systems, and networking infrastructure for large-scale AI deployments. Learn about the transformative impact of AI across industries and its role in driving economic growth, while understanding how collaborative efforts with the OCP community are shaping the future of accelerated computing.
Designing Data Centers for Tomorrow's AI Workloads - GPU Architecture and Rack-Scale Solutions
Open Compute Project via YouTube
Overview
Syllabus
Introduction
Nvidias contributions to OCP
Future of AI
Training at Scale
Outro
Taught by
Open Compute Project
Related Courses
-
Deploying the AI Factory - Transforming Data Centers with Accelerated Computing
-
Pushing Infrastructure Boundaries: Memory Requirements in AI Data Centers
-
Rapid Deployment of AI Solutions at Scale - Power, Cooling, and Networking Integration
-
Scalable AI-Ready Rack Cooling Solutions
-
The Exponential Demands AI Places on the Rack and Datacenter
-
Addressing Future Thermal Challenges Driven by AI - Data Center Cooling Solutions
