共查询到19条相似文献,搜索用时 78 毫秒
1.
根据VOD的特点开发了两种基于访问频率的替换算法:LFRU(least frequeney and recently used)和PLFU(period least frequency used)算法,它们都试图将访问频率大的视频数据保留在Cache中。LFRU算法结合了数据的访问频率和访问时间信息,对访问模式的变化具有一定的适应性。PLFU算法用周期法和预测法解决TLFU算法中的Cache“污染”问题。 相似文献
2.
3.
4.
随着网络上光盘资源的增长,光盘服务器成为光盘网络共享的重要技术,针对传统光盘服务器的缺点,实现了一个新的高性能光盘服务器CDS(CDServer),CDS系统采用两级Cache(客户端Cache加服务器Cache)的技术来提高系统的性能.客户端Cache根据光盘顺序访问特点,采用慢速增长快速下降的预取算法设计,即提高了系统的性能,也保证预取不命中时的响应时间,服务器Cache采用Hash算法同平衡二叉树相结合的两级组织结构,实现了Cache的快速查找.在详细介绍了CDS系统的两级Cache算法的同时,进行了相应的试验测试和性能分析. 相似文献
5.
作为提高CPU读取和存储数据的效率,弥补与主存之间存取速度差距的有效策略,CPU的缓存(Cache)充分利用其对数据使用的局部性原理,对最近或最常使用的数据进行暂存,对CPU的性能起着决定性作用.缓存的微架构正是决定缓存性能的关键性因素.然而,现代先进的CPU缓存都具备极为复杂的结构,存在多种策略、多种硬件算法和多个层级等不同维度的设计,从硬件上直接设计和论证不仅耗时而且成本很高,Cache微架构模拟器正是用软件方法对硬件微架构进行模拟和仿真.设计一款结构优良的缓存,对不同微架构进行评估,是一件具有深远意义的工作.本文从硬件结构出发,设计实现了一款多级、高可配、高可扩展的缓存微架构功能模拟器CMFSim(Cache microarchitecture functional simulator),实现了常见的缓存策略和硬件算法,可以进行给定配置下的缓存功能的模拟,从而分析配置参数与缓存性能间的关系. 相似文献
6.
在嵌入式处理器中,Cache的功耗所占的比重越来越大。为降低嵌入式系统中混合Cache的功耗,引入一种基于程序段的重构算法——PPBRA,并提出一种新的基于分类访问的可重构混合Cache结构,该方案能够根据不同程序段对Cache容量的需求,动态地分配混合Cache的指令路数和数据路数,还能够对混合Cache进行分类访问,过滤对不必要路的访问,从而实现降低混合Cache的功耗的目的。Mibench仿真结果表明,该方案在有效降低Cache功耗的同时,还能提高Cache的综合性能。 相似文献
7.
一种低功耗可重构Cache的重构算法 总被引:4,自引:0,他引:4
随着半导体技术的发展,芯片上的功率密度也逐渐增大,这使得功耗问题在芯片设计时越来越受到人们的关注.片上Cache是处理器芯片中的主要功耗源之一,采用低功耗Cache可有效降低处理器整体功耗.对低功耗Cache设计进行了研究.介绍了当前低功耗Cache设计的主要方法和一种低功耗可重构的数据Cache的体系结构及相应的重构算法.给出了一种新的重构算法——LoW-High Boundary(LHB)算法.实验表明LHB算法在性能和功耗上均优于原算法. 相似文献
8.
本文基于简单常见模式压缩编码设计了一种新颖的片内压缩Cache层次结构。在该结构中,L1数据Cache和L2Cache都以压缩格式保存数据,但具有不同的布局。其中,L1数据Cache的布局能触发部分Cache行预取,同时又能避免普通预取技术可能导致的Cache污染增加以及带宽浪费的现象,而且没有预取缓冲开销。实验结果表明,与传统Cache结构相比,本文的设计方案可以显著增加L1数据Cache和L2Cache的有效容量,并且不会增加L1数据Cache的访存延迟,对L1数据Cache平均能增加33%的有效容量,减少L1数据Cachhe失效率达21%,程序执行速度提高了13%。 相似文献
9.
嵌入式处理器的Cache结构研究 总被引:5,自引:0,他引:5
陈章龙 《小型微型计算机系统》2004,25(7):1204-1206
针对嵌入式处理嚣结构的特点,探讨虚拟Cache的结构、性能及实施方法等进行,讨论了Cache的锁定来改进Cache的循环淘汰置换算法的可行性,并对基于ARM架构的嵌入式处理器的Cache结构特点作了介绍。 相似文献
10.
循环分块是一种提高循环Cache命中率的循环变换技术,循环分块的大小是决定循环分块效率的关键因素,CME(cache miss equations)是一种精确分析程序中循环Cache命中率的数学模型,从CME理论模型出发,通过比较循环分块前后CME的变化,结合PADDING技术可以得出一个循环分块算法。实验表明,通过该算法计算出的块大小较之经典的LRW循环分块算法,在确保完全消除循环中数且引用数据访问Cache自冲突的同时,可以获得更大的分块,从而提高了循环分块的分块效率。 相似文献
11.
12.
黄明志 《电脑与微电子技术》2012,(11):7-9,12
阐述LRU算法的基本原理.提出在.NETFramework平台上基于LRU算法的高速缓存的设计思路.并给出一个通用、高性能、GB级、线程安全和支持泛型的LRUCache缓存类的具体实现方法。 相似文献
13.
14.
15.
设计了一个Cache数据一致性演示系统,能演示Cache在采用不同的映象规则与不同写策略时的动态读写过程. 相似文献
16.
Alan Jay Smith 《Performance Evaluation》1981,1(4):249-262
The file system, and the components of the computer system associated with it (disks, drums, channels, mass storage tapes and tape drives, controllers, I/O drivers, etc.) comprise a very substantial fraction of most computer systems; substantial in several aspects, including amount of operating system code, expense for components, physical size and effect on performance. In a comparison paper, we surveyed the traditional methods for optimizing the I/O system. We then examined disk and I/O system architecture in IBM type systems, and indicated shortcomings and future directions. In this paper we go one step further and summarize research by the author on two topics: cache disks and file migration. Cache disks are disks which have an associated cache which buffers recently used tracks of data. The case for cache disks is presented, and some of the issues are discussed. Parameter values for some aspects of the cache design are suggested. The second part of this paper summarizes the author's work on file migration, by which files are migrated between disk and mass storage as needed in order to effectively maintain on-line a much larger amount of information than the disks can hold. Some of the algorithms investigated are discussed, and the basic results are presented. 相似文献
17.
Alan Jay Smith 《Performance Evaluation》1981,1(3):249-262
The file system, and the components of the computer system associated with it (disks, drums, channels, mass storage, tapes and tape drives, controllers, I/O drivers, etc.) comprise a very substantial fraction of most computer systems; substantial in several aspects, including amount of operating system code, expense for components, physical size and effect on performance. In a companion paper, we surveyed the traditional methods for optimizing the I/O system. We then examined disk and I/O system architecture in IBM type systems, and indicated shortcomings and future directions. In this paper we go one step further and summarize research by the author on two topics: cache disks and file migration. Cache disks are disks which have an associated cache which buffers recently used tracks of data. The case for cache disks is presented, and some of the issues are discussed. Parameter values for some aspects of the cache design are suggested. The second part of this paper summarizes the author's work on file migration, by which files are migrated between disk and mass storage as needed in order to effectively maintain on-line a much larger amount of information than the disks can hold. Some of the algorithms investigated are discussed, and the basic results are presented. 相似文献
18.
嵌入式处理器中降低Cache缺失代价设计方法研究 总被引:2,自引:0,他引:2
以龙芯1号处理器为研究对象,探讨了嵌入式处理器中降低Cache缺失代价的设计方法.通过分析处理器的结构特征,本文实现了在关键字优先基础上一次缺失下命中的非阻塞数据Cache,可以将处理器平均性能提高3.9%,同时利用局部性原理,在关键字优先非阻塞数据Cache的基础上,本文提出了一种类非阻塞的指令Cache设计方法,可以降低指令Cache的缺失代价,以较小的实现代价进一步将处理器平均性能提高7.7%.通过本文的工作,可以同时降低指令Cache和数据Cache的缺失代价,处理器的平均性能提高了11.6%. 相似文献
19.
The finite length absolute minimum vectors of the Constant Modulus (CM) criterion with an arbitrary source and a single output linear regression model are shown to equal the Wiener filter or one of its phase shifted versions. The exact equations for these vectors are also provided and depend on the second and the fourth order statistical moments of the sample sequence only. These expressions describe the behavior of the Constant Modulus Algorithm (CMA) when it operates at the desired global minimum and expose key characteristics of the CM minimization approach. These characteristics include the existence conditions, the persistency of excitation, the computational complexity, the excess CM cost, the robustness to an additive noise, and the tolerance to the length variations. Moreover, several computer simulation experiments are included to validate our findings. 相似文献