期刊名称：International Journal of Mechatronics, Electrical and Computer Technology
印刷版ISSN：2305-0543
出版年度：2018
卷号：8
期号：29
页码：3959-3965
出版社：Austrian E-Journals of Universal Scientific Organization
摘要：Since 1990, power has been one of the major restrictions in computer systems. Due to lower energy
reserves, the importance of saving power is increasingly felt. Today this issue is much more
evident in computer science, especially computer programs. Further savings in power is more
acceptable in different software aspects because power saving in software aspects is more cost effective
than the hardware aspects. Power consumption is a major challenge in programs, thus code optimization
algorithms accelerate application's navigation and optimizes reduced power consumption. In the
applications, by reducing the consumed power, the processor will have free time and its waste of time
is prevented. For example, when a program runs in a CPU with slow clock, it consumes less energy than
is the case when it runs with higher clock speed. The purpose of this paper is to reduce power
consumption by applications. The power consumption has become a major challenge in computer
systems and knowledge and access to parts of the program that have higher power consumption is very
important because, by having access to these parts, it is possible to examine their improvement methods.
Here, genetic and K-means algorithms are combined and an algorithm is proposed for code optimization
that uses the loop interchange optimization technique which can reduce the power consumption by the
program. The implementation results indicate the reduced power consumption in implementation of
various matrix sizes. Program processing. The purpose of code optimization algorithm is to improve the
performance of programs to reduce their power consumption.
关键词：Code optimization algorithms; reducing power consumption; genetic algorithm; K;means
algorithm
