Parallel Real-Time Systems for Latency-Critical Applications

Jing Li
Washington University in St. Louis


Real-time systems need to provide quality of service guarantees for latency-critical applications that interact with humans and/or the physical environment. Examples span from cyber-physical systems, such as autonomous vehicles, avionic systems and robotics, to interactive cloud services, such as web search, stock trading, and online gaming. However, as parallel machines are becoming ubiquitous, we face challenges in designing real-time systems that can fully utilize the efficiencies of parallel computing platforms. In this talk, I will present some of my work that exploits the internal parallelism of latency-critical applications to design, analyze and implement parallel real-time systems for CPS and interactive cloud services. These systems exploit the untapped efficiencies in the parallel platforms and drastically improve the scalability and quality of service guarantees in widely used applications.