ENERGY EFFICIENT TASK SCHEDULING USING DVFS SCHEME
Keywords:
Task scheduling, Heterogeneous Earliest Finish Time, Dynamic voltage and frequency scalingAbstract
For redundant energy consumption, power optimization appears as a key performance parameter for the HPC (High Performance Computing) platform. Scheduling plays a vital role in attaining minimum make-span. The main focus of this paper is to minimize the energy consumption of application using DVFS. In this research, an algorithm named as is proposed for efficient energy to schedule DAG on heterogeneous graph G(V,E). The system performance is enhanced by reducing the energy consumption by considering computation energy of processors. DVFS with HEFT task scheduling algorithm is proposed. The proposed algorithm copes effectively and reduces energy consumption of applications. The work has generated heterogeneous graphs from homogenous graph models.
From the experiments, it is observed that when DVFS is used, the energy consumption is reduced by 11.77%
