Muhammad Shahbaz - Fast and Flexible Data-Center Networks

to

Location

Phillips Hall 233

Description

Abstract: Modern data-center networks need to be fast (high throughput and low latency) and flexible (programmable). Today, with the demise of Moore’s Law and the rise of big data, end-host servers are struggling to scale and meet the demands of current and emerging data-center workloads, such as group communication and machine-learning training. Switches in the network can make these workloads run orders of magnitude faster, by going beyond traditional packet forwarding to perform tasks such as in-network aggregation and multicast. The challenge, however, is how to enhance network flexibility without sacrificing performance. Exposing the right programming abstractions and hardware primitives for networks can enable more flexibility with negligible (or zero) loss in performance. In this talk, I will present systems and approaches (i.e., Pisces and Elmo) that let network programmers modify switch behavior using high-level languages (e.g., to encode state in packets efficiently) and new hardware primitives (i.e., to read, infer, and act on the encoded state), without compromising speed. Together, these features allow network operators to scale data-center workloads, including group communication, to millions of participants, while operating at line-rate. Bio: Muhammad Shahbaz is a postdoc in the Electrical Engineering Department at Stanford University. His research focuses on the design and development of domain-specific abstractions, compilers, and architectures for emerging workloads (including machine learning and self-driving networks). Shahbaz received his Ph.D. and M.A. in Computer Science from Princeton University and B.E. in Computer Engineering from the National University of Sciences and Technology (NUST). Before joining Princeton University, Shahbaz worked as a Research Assistant at Georgia Tech and the University of Cambridge, Computer Laboratory. Shahbaz has built open-source systems, including Pisces, SDX, and NetFPGA-10G, that are widely used in industry and academia. He received the Facebook Research Award, ACM SOSR Systems Award, APNet Best Paper Award, Internet2 Innovation Award, and Outstanding Graduate Teaching Assistant Award.