Real-time worst-case temperature analysis with temperature-dependent parameters

With the evolution of today’s semiconductor technology, chip temperature increases rapidly mainly due to the growth in power density. Therefore, for modern embedded real-time systems it is crucial to estimate maximal temperatures early in the design in order to avoid burnout and to guarantee that the system can meet its real-time constraints. This paper provides answers to a fundamental question: What is the worst-case peak temperature of a real-time embedded system under all feasible scenarios of task arrivals? A novel thermal-aware analytic framework is proposed that combines a general event/resource model based on network and real-time calculus with system thermal equations. This analysis framework has the capability to handle a broad range of uncertainties in terms of task execution times, task invocation periods, jitter in task arrivals, and resource availability. The considered model takes both dynamic and leakage power as well as thermal dependent conductivity into consideration. Thorough simulation experiments validate the theoretical results.