Keynote: Michael Franz (University of California, Irvine)
Fast and Furious: How the web got turbo charged just in time…
This is an award speech for the 2020 ACM Thacker Breakthrough in Computing Award.
Bio: Michael Franz is a Chancellor’s Professor in the Department of Computer Science at the University of California (UC), Irvine where he also directs the Secure Systems and Software Laboratory. His current research emphasis is in software systems, particularly focusing on compiler, virtual machine, and related system-level techniques for making software safer, or faster, or both.
Franz received a Doctor of Technical Sciences degree in Computer Science and a Diplomingenieur, Informatik-Ing. ETH degree, both from the Swiss Federal Institute of Technology (ETH Zurich). His honors include receiving a Humboldt Research Award from the Alexander von Humboldt Foundation, a National Science Foundation CAREER Award, and an IEEE Computer Society Technical Achievement Award. Franz is a Fellow of ACM, the Institute of Electrical and Electronics Engineers (IEEE), the American Association for the Advancement of Science (AAAS), and the International Federation for Information Processing (IFIP).
Keynote: Phillip Stanley-Marbell (University of Cambridge)
Enabling Trustworthy Autonomous Systems with Uncertainty-Tracking Computer Architectures
Autonomous systems such as self-driving cars, drones, and robots are rapidly-evolving technologies and will touch every aspect of human life. To be widely deployed, they must be trusted by humans. Today, they are not. One reason for the lack of trust is that the empirical data which drive autonomous systems (for example, from inertial sensors and cameras) are by nature uncertain, yet microprocessors today cannot track this uncertainty and its evolution in computations. This talk will present insights from research as well as commercial deployment of in-processor representations of uncertainty and an uncertainty-tracking computing platform which executes unmodified RISC-V binaries. The talk will highlight applications ranging from variational quantum algorithms and sensor data processing, to aerospace materials properties modeling and robotics. Unlike methods which require software to be rewritten in a domain-specific language or which may require extensive source-level changes, microarchitecture-level uncertainty tracking can achieve the same accuracy as Monte Carlo evaluation using approximately 2000-fold fewer instructions on average (and up to ~20k times fewer instructions in some cases) while requiring no source-level changes and with only minimal changes required to get uncertainty information into the microarchitecture.
Bio: Phillip Stanley-Marbell is an Associate Professor in the Department of Engineering at the University of Cambridge, UK (2017 to present), where he leads the Physical Computation Laboratory. He is the founder and CTO at Signaloid and is a Royal Academy of Engineering Enterprise Fellow (2022) and a Faculty Fellow at the Alan Turing Institute (2017 to 2021). Prior to moving to the UK in 2017, he was a researcher in the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT (2014–2017). From 2012 to 2014, he was with the Core OS organization at Apple (Cupertino, USA) where he led the development of new system components for iOS, macOS, and watchOS that enable on-device machine learning. Prior to Apple, he spent several years (2008–2012) as a permanent research staff member at IBM Research in Zürich, Switzerland. He completed his Ph.D. at Carnegie Mellon University (Pittsburgh, USA) in 2007, spending 2006–2008 at Technische Universiteit Eindhoven in the Netherlands. Before his Ph.D., he spent several summers as an intern and full-time engineer at Bell Labs (1995, 1996, 1999). His research focuses on investigating methods to use properties of physical systems to improve the efficiency of computation on data from nature. His research has led to several best paper nominations and awards (IEEE ESWEEK / Transactions on Embedded Computing Systems, ACM Computing Surveys), research highlights in the ACM’s flagship Communications of the ACM journal (CACM, 2021), as well as multiple long-form articles and profiles covering his research in the mainstream media (e.g., Fast Company 2019, Wired Magazine 2020).
Keynote: Tim Harris (Microsoft)
Hey, you got your distributed algorithm in my ML!
Bio: My current focus is multi-GPU inference and training of PyTorch models in the ONNX runtime system at Microsoft. More generally, my research interests span the stack encompassing distributed systems, language runtime systems, and operating systems, and with a particular emphasis on scalability and performance.
Prior to Microsoft, I was with AWS and worked on large-scale storage performance and data analytics with Amazon S3—for more on that, see talks from re:Invent, and the S3 Select feature. Further back, I led the Oracle Labs group in Cambridge, UK working on runtime systems for in-memory graph analytics, and the confluence of work on “big data” and ideas from high-performance computing. Before joining Oracle I had a previous stint with Microsoft (2004–2012), working on transactional memory and on the Barrelfish research operating system. I was on the faculty of the University of Cambridge Computer Laboratory (2000–2004) where I led the department’s research on concurrent data structures and contributed to the Xen virtual machine monitor project.