Design for Security

PAPER REFERENCES:

Part 1: Hardware Backdoors: http://www.cs.columbia.edu/~simha/preprint_oakland11.pdf
Part 2: SVF: http://www.cs.columbia.edu/~simha/preprint_isca12_svf.pdf
Part 3: Malware Detector: http://www.cs.columbia.edu/~simha/preprint_isca13_malware.pdf

Security is a full-system property. All components in a system "” including the hardware "” should be secure to create a secure system. However, hardware hasn't received much attention despite the fact that it can make or break security in fundamental ways. In this talk I will discuss how systems designed for security from hardware up offer unique advantages not available in current systems. I will illustrate the benefits through three case studies.

First, I will discuss how hardware itself can be created in a manner that provides assurance that its functionality has not been compromised due to design-time backdoors. Then, moving up the stack from hardware design to architectural design, I will discuss a new tool computer architects can use to reason about a class of vulnerabilities known as side channel leaks. I will show how this tool can be used to make performance-security tradeoffs if necessary. Finally I will describe the world's first hardware-only malware detector.
Simha Sethumadhavan is an Associate Professor of Computer Science at Columbia University. Simha's research interests are in hardware security and energy-efficient computing. He is the principal investigator on the DARPA SPARCHS project at Columbia University. The goal of the project is to discover how systems should be designed if security was a first order design requirement in addition to the traditional requirements like power/performance etc. He has been recognized with an Alfred P Sloan Fellowship (2013), NSF CAREER award (2011), two IEEE Micro "top pick" awards (2004, 2013) and a graduate teaching award (2006). He obtained his PhD from UT Austin in 2007.