TITLE: Efficient Graph Processing with Cross-Stack Optimization

ABSTRACT:

Graph analytics has emerged as an important way to understand the relationships between the heterogeneous types of data, allowing data analysts to draw valuable insights. However, graph processing poses key challenges for both architecture and runtime system. In this talk, I will demonstrate efficient graph processing at different scales with cross-stack optimization. First, I will present a distributed graph processing system using 3D graph partition to reduce communication. Next, I will present a out-of-core graph processing system using graph abstraction to accelerate convergence. Finally, I will discuss a graph processing accelerator based on the emerging ReRAM technology.

BIO:

Xuehai Qian is an assistant professor at the Ming Hsieh Department of Electrical Engineering and the Department of Computer Science at the University of Southern California. He received the Ph.D. from the Computer Science Department at University of Illinois at Urbana-Champaign working on parallel computer architecture. Prior to joining USC, he was a postdoctoral researcher at UC Berkeley. His recent research interests include system and architecture for graph processing, machine learning acceleration, NVM, and cloud system. He is the recipient of NSF CRII award and CAREER award.