Data Center Networking Systems

3.0 credits · Graduate

This is a graduate-level course that covers advanced topics in next-generation data center networking systems. It explores the intersection of computer networking, computer systems, and computer architecture, emphasizing how these traditionally separate domains interact to support large-scale, high-performance communication infrastructures in modern data centers. The course develops a principled understanding of end-to-end datacenter network stack design. Topics include datacenter network architectures (e.g., network topologies and communication protocols), modern networked applications and their communication patterns (e.g., machine learning, data analytics, microservices), advanced network interconnects (e.g., GPU interconnects, optical interconnects, memory interconnects), and programmable networking devices (network switches and network interface card). The goal of this course is to introduce fundamental networking design principles and equip students with a deep understanding of how the performance demands and communication patterns of modern applications shape the design of next-generation networking systems.