Cache替换策略的价值评估

发布时间：2018-07-10 17:05

本文选题：距离 + 替换策略评估　；参考：《中国科学技术大学》2017年硕士论文

【摘要】：Cpu容量小速度快,存储器容量大但是读取慢。随着科技的发展两者的速度差了几百倍,而Cpu访问主存的速度问题影响了计算机的速度,Cache作为一个中间的存储点被用来解决这个问题。当Cache的命中率足够高的时候,就相当于用Cache的速度来访问存储器,极大地提高了电脑的性能。为了提高Cache的命中率,产生了许多对Cache存储的策略,Cache的策略研究十分重要。但是任何一种策略对某种环境下相对于其他策略是有优势的。那么怎么衡量这些策略的优劣呢?同时对于一种新的数据怎么找到最适合它的Cache策略呢?本文首先总结了 Cache问题,然后基于距离概念和命中函数的定义,用理论分析和模拟的方法对各种替换策略进行评估,找到方法用来计算Cache块容量大小的价值,提出了一个新的算法,并预估这个算法的优劣。最后给出了一定数据下的最优策略,以及Cache大小对命中率的影响。
[Abstract]:CPU capacity is small and fast, memory capacity is large, but reading is slow. With the development of technology, the speed of the two is hundreds of times lower, and the speed of CPU accessing the main memory affects the speed of the computer. Cache is used as an intermediate storage point to solve this problem. When cache hit rate is high enough, it is equivalent to using cache speed to access memory, which greatly improves the performance of computer. In order to improve the hit rate of cache, it is very important to study the cache policy of cache storage. But any strategy has an advantage over other strategies in one environment. So how do you measure these strategies? At the same time, how to find the most suitable cache policy for a new data? In this paper, the cache problem is summarized, and then based on the definition of distance and hit function, various substitution strategies are evaluated by theoretical analysis and simulation, and the value of cache block capacity is calculated. A new algorithm is proposed and its advantages and disadvantages are estimated. Finally, the optimal strategy under certain data and the effect of cache size on hit rate are given.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP333

【相似文献】