期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2011
卷号：2
期号：4Ver1
出版社：Ayushmaan Technologies
摘要：The utility of cache and buffer is to reduce the time consumption when there is a data request. Due to the locality theory, the cache prefetches data for the future data request. Standing on the performance point of view, we need a mass cache to prefetch all data request. On the other hand, because of price and techniques reasons, the cache size is about dozens of mega bytes. With the capacity limit, a lot of cache management algorithms focus on two problems. Firstly which pages are to be prefetched into cache? Secondly which pages are to be swapped out of cache? Many cache management algorithms try to keep the pages with the greatest possibility of future access. The methodology of these algorithms is to maximize the possibility of cache hit. The cost of cache miss is expensive. Accordingly, new cache management algorithms are discussed in this paper.
