Self-Adjusting Networks for Optimized Datacenter Topology Design
Scalable Parallel Computing Lab, SPCL @ ETH Zurich via YouTube
Learn Excel & Financial Modeling the Way Finance Teams Actually Use Them
MIT Sloan: Lead AI Adoption Across Your Organization — Not Just Pilot It
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 40-minute conference talk by Stefan Schmid on self-adjusting networks, presented at SPCL_Bcast on October 27, 2022. Delve into the challenges of modern datacenter topology designs, including static, dynamic demand-oblivious, and dynamic demand-aware switches. Discover the concept of self-adjusting networks that optimize and match their traffic workload. Learn about information-theoretic metrics for quantifying communication traffic structure and achievable performance in demand-matching datacenter networks. Examine network design principles, identify open research challenges, and understand how self-adjusting networks and demand-aware graphs relate to classic optimization problems in theoretical computer science. The talk begins with a brief speaker introduction, followed by the main presentation.
Syllabus
Speaker Introduction
Talk
Taught by
Scalable Parallel Computing Lab, SPCL @ ETH Zurich