Jian Huang

Abstract Title

Photo of Professor Jiang Huang
Jian Huang

 

Towards Sustainable AI Infrastructure: Chips, Servers, and Racks 
This surge in AI has become the major source of the increasing energy consumption of data centers, leading to a significant increase in data center greenhouse gas emissions over the past decade. The energy efficiency of AI infrastructure is determined by multiple factors that include AI chips, server architecture, and rack organization. To ensure AI infrastructure to be developed and operated in a sustainable manner, we need to investigate possible solutions with a holistic approach, ranging from rack-scale data center deployment to low-level microarchitecture optimizations of AI chips. In this talk, I will present our recent research on sustainable AI infrastructures, including modular data center deployment with renewable energy, energy-aware operating system for server management, and AI chip innovation for improved energy efficiency.  

Biography

Dr. Jian Huang is an Associate Professor and Y. T. Lo Faculty Fellow in the ECE department at the University of Illinois at Urbana-Champaign. He received his Ph.D. in Computer Science at Georgia Tech in 2017. His research interests include computer systems and architecture, sustainable AI infrastructure, AI chips, memory/storage systems, and especially the intersections of them. He is leading the Systems Platform Research Group (Illinois PlatformX) to work on sustainable AI infrastructures. His research contributions have been published at top-tier computer architecture and systems conferences such as ISCA, MICRO, ASPLOS, OSDI, and SOSP. His work received USENIX Best Paper Award, MICRO Best Paper Runner Up, MICRO Best Artifact award, multiple IEEE Micro Top Picks (and Honorable Mentions), and Microsoft Research Outstanding Project Award. He also received the inaugural SIGMICRO Early Career Award, NSF CAREER Award, NSF CRII Award, Dean’s Award for Early Innovation, and multiple industry awards. He is the founder of the Workshop on Hot Topics in System Infrastructure (HotInfra).