Computer Architecture as a field is more than just the incremental pursuit of faster processors – it can include the search for fundamentally new approaches to computing that better serve the needs of humanity. One area where the gap between “better” and “faster” remains large is in supporting absolutely critical computations in the face of adversaries (e.g. handling finances, flying a plane, counting votes). Knowing that a computer system won’t do something (e.g. leak all secrets) despite the best efforts of a dedicated and diligent foe is something that has proven incredibly difficult for some very good reasons. Drawing insights from programing languages, machine learning, circuit design, information theory, computer security, and formal methods we describe how computers can support trustworthiness in much deeper and more verifiable new ways than previously thought possible. These and other approaches are a step towards a future where we more fully distinguish between "intuitive" and "deliberative" styles of computing yet where we can enjoy the benefits of both working together.
Tim Sherwood is a Professor of Computer Science and the Associate Vice Chancellor for Research at UC Santa Barbara. He is a 9-time winner of the IEEE Micro Top Pick Award, a co-founder of the hardware security company Tortuga Logic, and the 2016 ACM SIGARCH Maurice Wilkes Awardee "for contributions to novel program analysis advancing architectural modeling and security". A proud product of public schooling from start to finish, he did his Ph.D. work at UCSD and B.S. at UC Davis.
This Distinguished Lecture is part of the Illinois Computer Science Speaker Series.