Energy-efficient Scheduling Algorithm for End-to-end Tasks in Heterogeneous Multi-core System

• Main Authors: Chun-Hsiung Haung, 黃群雄
• Other Authors: Chin-Fu Kuo
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/23969080807466973330

碩士 === 國立高雄大學 === 資訊工程學系碩士班 === 100 === Due to the complicated digital signal processing of multimedia systems, the number of cores on a heterogeneous multi-core platform progressively increases. But the hardware cost and the energy consume also progressively increase. Therefore, in this paper, we address the problem of reducing the cost of hardware, and the energy saving on such a system. First, we proposed an offline mechanism to minimize the number of processors needed by a task set and allocated the tasks to the processor. Second, with the dynamic voltage/frequency scaling we proposed an online EDF-based scheduling mechanism to dynamically adjust the processor frequency for jobs of the tasks. Finally, by the experimental results we confirm that the offline mechanism can indeed reduce the cost of hardware, and the online mechanism can indeed reduce the energy consumption.