Design-time energy optimization for asymmetric multiprocessor system-on-chip
Yonghee Yun, Young Hwan Kim · 2016
This paper proposes a static task mapping algorithm for energy-efficient embedded system design. The proposed algorithm is based on a genetic algorithm, which generates a set of solution candidates and evolves the candidates using genetic operations. During the evolvement, the proposed approach evaluates the time slack of each candidate, and then updates the candidates using a novel adaptive generation method. Experimental results show that the proposed method reduced the energy consumption by up to 16.1% and 34.1% compared to the existing methods and outperformed them in most cases.