文摘
We have demonstrated that depending on the busy/idle percentage of a program and also the overall execution time, how the proposed DVFS mechanism should be modified to achieve the maximum energy savings without paying significant timing overhead. In this paper we have proposed a complier-based DVFS methodology to improve energy dissipation of a multi-core processing platform. The proposed compiler-based DVFS mechanism performs better than the existing history- or profile-based methods that predict the future program behavior depending on the past statistics. Although in some of the compute-intensive programs our algorithm does not save as much energy as history-based algorithm, it always gets a lower energy-delay product, even lower than profile-based algorithm in most cases.