Power minimization techniques on distributed real-time systems by global and local slack management

摘要

Recently, a static power management with parallelism (P-SPM) technique has been proposed to reduce the energy consumption of distributed systems to execute a set of real-time dependent tasks (Mishra et al., 2003). The authors claimed that the proposed P-SPM outperforms other known methods in energy reduction. However, how to take advantage of the local static slack for further energy optimization remains as an open problem. In this paper, we propose the static power management with proportional distribution and parallelism scheme (PDP-SPM) that not only answers this open problem, but also exploits the parallelism. Simulations on task graphs derived for DSP applications and TGFF benchmark suite suggest that PDP-SPM achieves 64% energy saving over the system without power management, and 15% over the P-SPM scheme.