The Quest for Efficient and Trustworthy Systems
Baris Kasikci, University of Michigan
The current growth of the software ecosystem is faster than ever before. Software systems are increasingly more complex and consist of deep stacks. Software’s soaring complexity is causing a shift to a more heterogeneous hardware landscape, comprising CPUs, GPUs, FPGAs, fast networks, and denser memory technologies. For the foreseeable future, improving the efficiency of computer systems will be crucial to enable society’s growing reliance on feature-rich software. Alas, trustworthiness in this complex ecosystem is often an afterthought. Consequently, software and hardware have been plagued with bugs that cause data loss, security vulnerabilities, and failures of critical infrastructure. The estimated cost of poor software quality in the US alone was above 2 trillion USD in 2020. Building systems that are simultaneously efficient (i.e., deliver high performance at scale, leverage heterogeneous resources) and trustworthy (i.e., contain fewer bugs, achieve greater security, and have verified properties) is extremely challenging. In this talk, I will give an overview of the research in my lab that focuses on developing techniques for building systems that are simultaneously efficient and trustworthy, with a focus on real-world technical and societal impact.
Bio: Baris Kasikci is an assistant professor in the Electrical Engineering and Computer Science Department at the University of Michigan. His research is centered around building efficient and trustworthy computer systems. His group has built a number of techniques to improve the efficiency of datacenter applications, provide systems support for heterogeneous platforms, verify properties of complex distributed systems, detect, analyze, and fix failures, and improve the security of modern hardware. Previously, Baris was a researcher in the Systems and Networking Group at Microsoft Research Cambridge, UK. He completed his Ph.D. in Computer Science at EPFL. He also held roles at Intel, VMware, and Siemens. He is the recipient of an NSF CAREER award, a Microsoft Research Faculty Fellowship, a VMware Early Career Grant, an Intel Rising Star Award, a Google Faculty Award, multiple Google and Intel research awards, a Jay Lepreau Best Paper Award at OSDI, IEEE MICRO Top Picks Award, multiple IEEE MICRO Top Picks Honorable Mentions, a VMware fellowship, the Roger Needham Ph.D. Award for the best Ph.D. thesis in computer systems in Europe, and the Patrick Denantes Memorial Prize for best Ph.D. thesis in the Department of Information and Communication Sciences at EPFL. More details can be found on his webpage https://web.eecs.umich.edu/~barisk/.