ABSTRACT
This chapter focuses on dynamic power management (DPM) policies, including those targeting dynamic voltage scaling (DVS). DPM encompasses a set of policies that achieve energy-efficient computation by selectively turning off or reducing the performance of system components when they are idle or partially unexploited. The fundamental premise for the applicability of DPM is that systems and their components experience nonuniform workloads during operation time. The various voltage scheduling methods in DVS-enabled systems differ by when to adjust frequency and voltage, how to estimate slacks, and how to distribute these to waiting tasks. DVS scheduling schemes are classified by voltage scheduling granularity and fall into two categories: inter-task DVS algorithms and intra-task DVS algorithms. DVS techniques for hard-real-time systems enhance the traditional earliest deadline first or rate monotonic scheduling to exploit slack time, which is used to adjust voltage and frequency of voltage-scalable components.
