孙玉强, 王文闻, 巢碧霞, 顾玉宛. 基于预取的Cache替换策略[J]. 微电子学与计算机, 2017, 34(1): 85-89, 94.
引用本文: 孙玉强, 王文闻, 巢碧霞, 顾玉宛. 基于预取的Cache替换策略[J]. 微电子学与计算机, 2017, 34(1): 85-89, 94.
SUN Yu-qiang, WANG Wen-wen, CHAO Bi-xia, GU Yu-wan. Cache Replacement Policy Based on Prefetch[J]. Microelectronics & Computer, 2017, 34(1): 85-89, 94.
Citation: SUN Yu-qiang, WANG Wen-wen, CHAO Bi-xia, GU Yu-wan. Cache Replacement Policy Based on Prefetch[J]. Microelectronics & Computer, 2017, 34(1): 85-89, 94.

基于预取的Cache替换策略

Cache Replacement Policy Based on Prefetch

  • 摘要: 在页式虚拟存储管理系统中, 合理的Cache替换策略可以减少CPU等待时间, 提升Cache命中率和系统性能.许多经典Cache置换算法的依据都是页面的访问次数或访问时间, 也有一些是两者兼顾.对此介绍了一种基于预取的Cache替换策略, 主要做出了以下2点优化: (1)一次Cache置换淘汰多个页面; (2)Cache首次调入页面由"拉取"模式改为"推送"模式.模拟实验结果显示, 该方法有效提高了Cache命中率, 提升了系统性能.

     

    Abstract: In page-virtual storage management system, So reasonable Cache replacement policy can reduce the waiting time of CPU and improve system performance. Many typical Cache replacement policies arebased on the accessing frequency or time of the page and some policiesdo both.This article will describe a Cache replacement strategy which is based on Prefetch. We mainly make the following two optimization: (1) eliminate multiple pages in one Cache page replacement; (2)change the distribution model of the first batch of tasks from the pull model to the push model. Simulation experiment shows that this method can effectively improve the hit rate of Cache.

     

/

返回文章
返回