Date: Friday, December 5
Time: 12-1pm
Location: Soda 510
Title: Building reliable and efficient kernels for the next generation workloads
Abstract: The ever-growing demand for massive data processing and increasing application complexity has resulted in a push towards modular, fine-grained programming and deployment paradigms, such as microservices and serverless computing. These paradigms defy the core design principles of traditional operating systems, hypervisors, and runtimes. My work tackles the reliability and efficiency challenges posed by these paradigms by rethinking our software foundations to make them support data-intensive, short-lived, and highly parallel computations.
Bio: Pedro Fonseca is an Associate Professor in the Department of Computer Science at Purdue University, where he leads the Reliable and Secure Systems lab. He works at the intersection of systems, security, architecture, and programming languages, and his recent work focuses on building a reliable and efficient software stack for emerging programming and deployment paradigms. Pedro is the recipient of an NSF CAREER Award, a Google Faculty Research Award, and a Google Research Scholar Award. He routinely transfers technology to industry and has worked with open-source developers to confirm and fix hundreds of bugs found using automated systems. Before Purdue, he graduated from MPI-SWS and the University of Saarland and was a postdoc at the University of Washington.