Monday February 18, 2019

Speaker: Srini Devadas (MIT)
Title: Towards Secure High-Performance Computer Architectures
Abstract: Recent work has shown that architectural isolation can be violated through software side channel attacks that exploit microarchitectural performance optimizations such as speculation to leak secrets. While turning off microarchitectural optimizations can preclude some classes of attacks, we argue that performance and security do not have be in conflict, provided processors are designed with security in mind.

We advocate designing processors to support enclaves, which are processes with an associated security policy, and espouse a principled hardware/software co-design approach to eliminating entire attack surfaces relevant to the security policy through microarchitectural isolation, rather than plugging attack-specific privacy leaks. As a case study of this approach, we describe both in-order and speculative processor designs that offer strong provable isolation of software modules running concurrently and sharing resources, even when large parts of the operating system are compromised. Open-source implementations of these processors will allow security properties to be independently verified. Finally, we describe the current limitations of this approach and future opportunities for research.
Bio: Srini Devadas is the Webster Professor of EECS at MIT where he has been on the faculty since 1988. His current research interests are in computer security, computer architecture and applied cryptography. Devadas received the 2015 ACM/IEEE Richard Newton award, the 2017 IEEE W. Wallace McDowell award and the 2018 IEEE Charles A. Desoer award for his research in secure hardware. He is a Fellow of the ACM and IEEE. Devadas is the author of "Programming for the Puzzled" (MIT Press, 2017), a book that builds a bridge between the recreational world of algorithmic puzzles and the pragmatic world of computer programming, teaching readers to program while solving puzzles. He is a MacVicar Faculty Fellow, an Everett Moore Baker and a Bose award recipient, considered MIT's highest teaching honors.

Tuesday February 19, 2019

Speaker: Karin Strauss (Microsoft Research)
Title:When Moore met Feynman: Ultra-dense data storage and extreme parallelism with electronic-molecular systems
Abstract: Sustaining Moore's law is an increasingly challenging proposition. This talk will cover an alternative approach: going directly to the molecular level, as suggested by Feynman in his famous lecture, "There's Plenty of Room at the Bottom." Although we have yet to achieve scalable, general-purpose molecular computation, there are areas of IT in which a molecular approach shows growing promise. In this talk, I will explain how molecules, specifically synthetic DNA, can store digital data and perform certain types of special-purpose computation by leveraging tools already developed by the biotechnology industry. I will also discuss the architectural implications of molecular storage and processing systems and advocate for hybrid electronic-molecular systems as potential solutions to difficult computational problems, such as large-scale similarity search.
Bio: Karin Strauss is a Principal Researcher at Microsoft Corporation and an Affiliate Professor at the University of Washington. She co-leads the Molecular Information System Laboratory with Luis Ceze, working on using molecules, currently DNA, to benefit the IT industry. Her background is in computer architecture, systems, and most recently biology. Her research interests include emerging storage technologies, scaling of computation and storage, and special-purpose accelerators. Selected as one of the "100 Most Creative People in Business in 2016" by Fast Company Magazine, she got her PhD from the Department of Computer Science at the University of Illinois, Urbana-Champaign in 2007.

Wednesday February 20, 2019

Speaker: Michael O'Boyle (University of Edinburgh)
Title: Rethinking Compilation in a Heterogeneous World
Abstract: Moore's Law has been the main driver behind the extraordinary success of computer systems. However, with the technology roadmap showing a decline in transistor scaling and hence the demise of Moore's law, computer systems will be increasingly specialised and diverse. The consistent ISA contract is beginning to break down. As it stands, software will simply not fit. Current compiler technology, whose role is to map software to the underlying hardware is incapable of doing this. This looming crisis requires a fundamental rethink of how we design, program and use heterogeneous systems. This talk examines new ways of tackling heterogeneity so that, rather than deny and fear the end of Moore's law, we embrace and exploit it.
Bio: Michael O'Boyle is a Professor of computer science at the University of Edinburgh. He is best known for his work in incorporating machine learning into compilation and parallelization, automating the design and construction of optimizing technology. He has published over 100 papers and received three best paper awards. He was presented with the ACM CGO Test of Time award in 2017. He is a founding member of HiPEAC, the Director of the ARM Research Centre of Excellence at Edinburgh and Director of the EPSRC Centre for Doctoral Training in Pervasive Parallelism. He is a senior EPSRC Research Fellow and a Fellow of the BCS.