Learn about the critical role of short-reach optics in scaling artificial intelligence processing units (xPUs) through this 19-minute technical presentation from Innolight's Ryan Yu at the Open Compute Project. Explore the technological challenges and solutions for implementing optical interconnects in AI computing infrastructure, focusing on how short-reach optical technologies enable efficient data transmission and processing in large-scale AI systems. Gain insights into the latest developments in optical networking specifically designed to meet the demanding requirements of modern AI accelerators and processing units.
Overview
Syllabus
Short Reach Optics for AI xPU Scale up
Taught by
Open Compute Project
Related Courses
-
Linear Optics for AI Fabric Scale-Up and Scale-Out
-
VCSELs in Co-Packaged Optics for Short-Range Applications
-
Steering Copper Architectures - Opportunities for Short-Reach Pluggable Optics
-
Energy-Efficient High Bandwidth Optical Interconnect Solutions for Generative AI
-
Optical Interconnects for Large-Scale AI Clusters - A Meta Perspective
-
Efficient Datacenter Scale AI Acceleration Using 3D Optics - Reducing Costs and Power Consumption
