Periodic thermal management for hard real-time systems

Abstract

Due to growing power density, on-chip temperature increases rapidly, which has hampered the reliability and performance of modern real-time systems. This paper studies how to minimize the peak temperature for hard real-time systems under hard real-time constraints with periodic thermal management. A closed-form representation of the peak temperature for such a periodic scheme is derived to tackle this problem. Based on this closed-form and the arrival curve model which is used to model the system workload, two approaches that can derive periodic thermal management are proposed to minimize the peak temperature for a given event stream with a trade-off between complexity and accuracy. Case studies show that our approaches can achieve similar or better level of peak temperature but with two or three orders of magnitude lower computation expense compared to previous work.