Elbrus处理器缓存中数据压缩机制的设计

Aleksey S. Kozhin, A. V. Surchenko
{"title":"Elbrus处理器缓存中数据压缩机制的设计","authors":"Aleksey S. Kozhin, A. V. Surchenko","doi":"10.1109/EnT50437.2020.9431291","DOIUrl":null,"url":null,"abstract":"Higher capacity of cache memory increases Hit Rate at the cost of power consumption and die area. Hardware data compression in cache allows increasing its effective capacity without significant negative effects on these characteristics. This work discusses the realization of data compression in L3 cache memory of modern processors with “Elbrus” architecture. One of the most efficient and at the same time fast compression algorithms is BDI (Base-Delta-Immediate) algorithm. Its modification, called BDI*-HL (Base-Delta-Immediate Modified, Half-Line), is selected for implementation. To achieve an increase in effective capacity, the principles of joint placement of compressed cache lines are introduced and implemented. L3 cache memory is modified to correctly work with compressed lines. A testing bench is developed to evaluate the effectiveness of the designed mechanism. The developed solution does not require any significant changes in the structure of cache memory, does not affect the principles of eviction policy and the coherency protocol and does not change values of working frequency and cache die area.","PeriodicalId":129694,"journal":{"name":"2020 International Conference Engineering and Telecommunication (En&T)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Design of Data Compression Mechanism in Cache Memory of Elbrus Processors\",\"authors\":\"Aleksey S. Kozhin, A. V. Surchenko\",\"doi\":\"10.1109/EnT50437.2020.9431291\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Higher capacity of cache memory increases Hit Rate at the cost of power consumption and die area. Hardware data compression in cache allows increasing its effective capacity without significant negative effects on these characteristics. This work discusses the realization of data compression in L3 cache memory of modern processors with “Elbrus” architecture. One of the most efficient and at the same time fast compression algorithms is BDI (Base-Delta-Immediate) algorithm. Its modification, called BDI*-HL (Base-Delta-Immediate Modified, Half-Line), is selected for implementation. To achieve an increase in effective capacity, the principles of joint placement of compressed cache lines are introduced and implemented. L3 cache memory is modified to correctly work with compressed lines. A testing bench is developed to evaluate the effectiveness of the designed mechanism. The developed solution does not require any significant changes in the structure of cache memory, does not affect the principles of eviction policy and the coherency protocol and does not change values of working frequency and cache die area.\",\"PeriodicalId\":129694,\"journal\":{\"name\":\"2020 International Conference Engineering and Telecommunication (En&T)\",\"volume\":\"15 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-11-25\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2020 International Conference Engineering and Telecommunication (En&T)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/EnT50437.2020.9431291\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 International Conference Engineering and Telecommunication (En&T)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/EnT50437.2020.9431291","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1

摘要

高容量的高速缓存提高了命中率,但代价是功耗和芯片面积。缓存中的硬件数据压缩允许增加其有效容量,而不会对这些特性产生显著的负面影响。本文讨论了采用“Elbrus”架构的现代处理器在L3高速缓存中的数据压缩实现。BDI (Base-Delta-Immediate)算法是目前最有效、最快速的压缩算法之一。它的修改,称为BDI*-HL (Base-Delta-Immediate Modified, Half-Line),被选择用于实现。为了实现有效容量的增加,介绍并实现了压缩缓存线的联合放置原则。L3缓存被修改为正确地使用压缩行。为评价所设计机构的有效性,建立了试验台。开发的解决方案不需要对缓存结构进行任何重大更改,不影响驱逐策略和一致性协议的原则,也不改变工作频率和缓存死区值。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Design of Data Compression Mechanism in Cache Memory of Elbrus Processors
Higher capacity of cache memory increases Hit Rate at the cost of power consumption and die area. Hardware data compression in cache allows increasing its effective capacity without significant negative effects on these characteristics. This work discusses the realization of data compression in L3 cache memory of modern processors with “Elbrus” architecture. One of the most efficient and at the same time fast compression algorithms is BDI (Base-Delta-Immediate) algorithm. Its modification, called BDI*-HL (Base-Delta-Immediate Modified, Half-Line), is selected for implementation. To achieve an increase in effective capacity, the principles of joint placement of compressed cache lines are introduced and implemented. L3 cache memory is modified to correctly work with compressed lines. A testing bench is developed to evaluate the effectiveness of the designed mechanism. The developed solution does not require any significant changes in the structure of cache memory, does not affect the principles of eviction policy and the coherency protocol and does not change values of working frequency and cache die area.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信