The University of British Columbia

January 15, 2020 | Graduate Student Achievements, Student Achievements

ECE PhD Alum Wins SPEC Distinguished Dissertation Award!

ECE PhD Alum Dr. Guanpeng (Justin) Li (left) and PhD thesis advisor Dr. Karthik Pattabiraman (right).

Recently graduated ECE PhD student Guanpeng (Justin) Li won a distinguished dissertation award for his PhD thesis, titled Understanding and Modeling Error Propagation in Programs. Li was awarded the 2019 Kaivalya Dixit Distinguished Dissertation Award by the Standard Performance Evaluation Corporation (SPEC) Research group, recognizing outstanding doctoral dissertations in the field of computer benchmarking, performance evaluation, and experimental system analysis in general. Nominated dissertations are evaluated in terms of scientific originality, scientific significance, practical relevance, impact, and quality of the presentation. Only one dissertation is awarded each year (since 2011). Dr. Li’s PhD thesis advisor was Dr. Karthik Pattabiraman.

The selection committee was especially impressed with the intellectual merit, potential impact, quality, and clarity of Dr. Li’s dissertation. Li’s thesis is in the area of dependable computer systems, or computer systems that must continue to work despite faults in their operation. Such systems are used in a wide variety of domains such as aerospace, banking, healthcare and most recently, in autonomous vehicles. Because of their critical nature, significant effort is expended in building and maintaining these systems.

Li’s thesis discusses how error propagation is often the leading cause of catastrophic system failures, and hence must be mitigated. He first characterizes error propagation in programs that lead to different types of failures, proposing both empirical and analytical approaches to identify and mitigate error propagation without expensive fault injections. His key observation is that only a small fraction of states are responsible for almost all error propagation in programs, and the propagation falls into identifiable patterns which can be modeled efficiently. The proposed techniques are nearly as close as fault injection approaches in measuring failure rates of programs, and orders of magnitude faster than fault injections. This allows developers to build low-cost fault-tolerant applications in an extremely efficient manner.

Dr. Pattabiraman, who was Dr Li’s PhD advisor says, “The main strength of Justin’s dissertation is that it combines theoretical methods with rigorous experimental validation, and hence has significant practical impact. His results have been adopted by leading research labs such as IBM, Nvidia, and Los Alamos National Labs. In addition, his research has spawned many follow up projects at other universities. He has also published his work at top-tier conferences in our area such as DSN, SC, and ISSRE.“

Li’s award will be presented at the Annual ICPE International Conference on Performance Engineering (ICPE), in Canada in April 2020. Dr Li also received the NSERC Post-Doctoral Fellowship (PDF), and is currently a post-doctoral researcher at the University of Illinois at Urbana Champaign (UIUC). He will soon join the University of Iowa’s Computer Science dept as a tenure-track assistant professor.

Learn more about Li’s winning dissertation here.

News article: https://research.spec.org/news/single-view/article/winner-of-spec-kaival...