This talk will be live-streamed and is open to the public. To receive Zoom meeting information, send an email request to: [email protected]

Secure multi-party computation (MPC) allows multiple parties, each holding private data that they are unwilling to share, to collaborate to perform computations over their data, while revealing nothing other than the result. The theory behind secure computation is already 35 years old, but in the last 15 years researchers in the field have demonstrated tremendous improvement in performance. Today, a confluence of factors are driving an increased interest in secure computation, both in the public domain and in industry, including the advancements in MPC mentioned above, the explosion in the volume of private user data being generated, and the advancements in machine learning that are being deployed to leverage that private data.

In this talk, I will give an overview of several of my recent results that has helped scale the performance of secure computation. In the first half of the talk, I will focus on the setting where we have large volumes of data and a secure computation that is out-source to a small number of servers — the Few-PC setting. I will focus on a security relaxation that allows us to improve communication costs asymptotically and concretely for a large class of computations. In the second half of the talk, I will describe several results that allow us to scale MPC to thousands or even millions of parties, by demonstrating a protocol that requires less communication and computation from each participant as the number of participants grows.

In the real world, people join humanity, grow up trusting very different quorums from one another, and they may change their quorums with time. In technical terms we could call this an open permissionless system with mutable local quorums.

One might expect such a system to be a jumbled disorder — and yet, somehow, it self-organises into uniform areas of fairly stable consensus. One might almost suspect there could be some deep mathematical principles involved!

In this talk I will discuss consensus in a permissionless open system with mutable local quorums, and show how a topological view of the system gives a clean mathematical analysis with surprisingly good explanatory power.

Increasing set of functionalities, network interoperability, and system design complexity have introduced easily exploitable security vulnerabilities in cyber-physical systems (CPS). Furthermore, the tight interaction between information technology and physical world, as well as the rising levels of autonomy, make these systems vulnerable to attacks beyond the standard cyber-attacks; relying exclusively on conventional security techniques may be unfeasible due to resource-constraints and long system lifetime. Consequently, there is a need to change the way we reason about security in CPS, and start designing platform-aware attack-resilient components and architectures capable providing strong safety and performance guarantees even under attack. In this talk, I will present research challenges and our recent efforts in this domain, starting from cyber-physical security techniques that (a) capture effects of attacks on system performance, (b) introduce attack-resilience into components at each level of the autonomy stack, and (c) enable mapping of the desired Quality-of-Control (QoC) under attack guarantees into real-time platform requirements in way that supports design-time tradeoffs between the QoC under attack and security-related overhead. For systems with varying levels of autonomy and human interaction, I will also show how we can exploit human power of inductive reasoning and the ability to provide context, to improve the overall security guarantees. Finally, I will present how we can capture security-related CPS specification as hyperproperties, and introduce statistical model checking methods to check such specifications.

The UTxO (unspent transaction output) model is the underlying data structure of Bitcoin, which has since been extended to the Extended UTxO model.

It exists in code, but what does it mean? I will give a novel mathematical model based on some strikingly simple type equations which -- for me at least -- make it easier to see what is structurally going on. I will describe how the equations can be used to obtain two further models, one which is more abstract and one which is more concrete:

• A universal algebra axiomatisation, exhibiting Blockchain as an algebraic structure and so implying a general mathematics and specific testable properties against which to verify an EUTxO implementation, and

• An executable Haskell reference implementation

The rise of machine learning (ML) and artificial intelligence (AI) have instigated significant interests in developing domain specific integrated circuits and architectures that can support the computational demands of AI. These chips are commonly known as AI chips. AI chips, both for training and inferencing, have emerged out to be new areas of research and commercialization in semiconductor industry. Particularly, increasing emphasis has been placed on developing low-power AI chips for edge applications where AI can be local, distributed, and seamlessly integrated in sensors or other internet of things (IoT) devices for building naturally-autonomous systems. On the other-hand, globalization of semiconductor ecosystem has led to an era where achieving dominance on Microelectronics supply chain has become as exciting as the famous series “Game of Thrones” for not only the companies but also the countries. Furthermore, the end of Moore’s Law of Scaling of traditional semiconductor transistors has added the climax to this situation as the development and manufacturing of advanced process nodes for the next generation of logic and memory devices get expensive driving most semiconductor industry in the US to be fabless. Interestingly, advanced process nodes such as FINFETS, 3D Transistors, RRAMs/Memristors, and 3D/Monolithic integration and packaging approaches provide unsurpassed opportunities to achieve the much desired On-Chip AI along with the primitives for implementing Hardware Security, and Trust into the integrated circuits. This talk will review some of these technologies and our research efforts in these areas.