内存与片上渗透缓存之间数据迁移的理论分析

doi:10.11959/j.issn.1000-436x.2021157

Abstract

Abstract:

To improve the computer processor’s memory access efficiency, increase hit rates, reduce data fetching latency, the data and instructors that had close relation with each other could build the just-in-time locality environment on processor’s cache.To build such an environment, the relations between the data and instructors should be studied while them being immigrated into on-chip cache.The research results show that these relations could be well kept when the data and instructors are moved.The simulation also shows that the percolation cache can help the processor core to increase its hit rates.These research results give a new way to build the just-in-time locality environment on the percolation cache.

Key words: data immigration on chip, just-in-time locality, percolation

CLC Number:

TP302.1

Jiuchuan HU, Dongrui FAN, Jiancong CHENG, Long YAN, Xiaochun YE, Lingzhi LI, Liangyi WAN, Haibin ZHONG. Theoretical analysis for the data immigration between memory and processor percolation cache[J]. Journal on Communications, 2021, 42(8): 217-225.

Figures/Tables 11

References 16

[1]	胡九川, 范东睿, 李丹萍, ,等. 一种支持数据渗透迁移的片上缓存模型研究[J]. 北京交通大学学报, 2017,41(5): 1-9.
	HU J C , FAN D R , LI D P ,et al. An on-chip cache model research on supporting data permeation and migration[J]. Journal of Beijing Jiaotong University, 2016,41(5): 1-9.
[2]	TAN G M , SUN N H , GAO G R . Improving performance of dynamic programming via parallelism and locality on multicore architectures[J]. IEEE Transactions on Parallel and Distributed Systems, 2009,20(2): 261-274.
[3]	GARCIA E , OROZCO D , KHAN R ,et al. A dynamic schema to increase performance in many-core architectures through percolation operations[C]// 20th Annual International Conference on High Performance Computing. Piscataway:IEEE Press, 2013: 276-285.
[4]	李丹萍 . 单核处理器片上渗透数据调配方法研究[D]. 北京:北京交通大学, 2016.
	LI D P . Research on allocation method of penetration data on single-core processor[D]. Beijing:Beijing Jiaotong University, 2016.
[5]	FAN D R , YUAN N , ZHANG J C ,et al. Godson-T:an efficient many-core architecture for parallel program executions[J]. Journal of Computer Science and Technology, 2009,24(6): 1061-1073.
[6]	ZHENG T H , ZHU H S , EREZ M . SIPT:speculatively indexed,physically tagged caches[C]// 2018 IEEE International Symposium on High Performance Computer Architecture. Piscataway:IEEE Press, 2018: 118-130.
[7]	CRUZ E H M , DIENER M , ALVES M A Z ,et al. LAPT:a locality-aware page table for thread and data mapping[J]. Parallel Computing, 2016,54: 59-71.
[8]	BHATTI M K , OZ I , AMIN S ,et al. Locality-aware task scheduling for homogeneous parallel computing systems[J]. Computing, 2018,100(6): 557-595.
[9]	王子聪, 陈小文, 郭阳 . 片上多核处理器Cache访问均衡性研究[J]. 计算机学报, 2019,42(11): 2403-2416.
	WANG Z C , CHEN X W , GUO Y . Research on cache access equalization in chip multi-processor[J]. Chinese Journal of Computers, 2019,42(11): 2403-2416.
[10]	GAUR J , CHAUDHURI M , RAMACHANDRAN P ,et al. Near-optimal access partitioning for memory hierarchies with multiple heterogeneous bandwidth sources[C]// 2017 IEEE International Symposium on High Performance Computer Architecture. Piscataway:IEEE Press, 2017: 13-24.
[11]	EL-SAYED N , MUKKARA A , TSAI P A ,et al. KPart:a hybrid cache partitioning-sharing technique for commodity multicores[C]// 2018 IEEE International Symposium on High Performance Computer Architecture. Piscataway:IEEE Press, 2018: 104-117.
[12]	NORI A V , GAUR J , RAI S ,et al. Criticality aware tiered cache hierarchy:a fundamental relook at multi-level cache hierarchies[C]// 2018 ACM/IEEE 45th Annual International Symposium on Computer Architecture. Piscataway:IEEE Press, 2018: 96-109.
[13]	YI L , SHAN G B , LIU S ,et al. High-performance processor design based on 3D on-chip cache[J]. Microprocessors and Microsystems, 2016,47: 486-490.
[14]	胡伟武, 陈云霁, 肖俊华 . 计算机体系结构[M]. 北京: 清华大学出版社, 2011.
	HU W W , CHEN Y J , XIAO J H . Computer architecture[M]. Beijing: Tsinghua University Press, 2011.
[15]	熊金城 . 点集拓扑讲义,第4版[M]. 北京: 高等教育出版社, 1993.
	XIONG J C . Point set topology lecture[M]. 4th edition,Beijing: Higher Education Press, 1993.
[16]	PATTERSON A , HENNESSY L . Computer organization and design:the hordware/software interface[M]. 5th edition,Morgan Kaufmann: Morgan Kaufmann Publisher, 2014.

Metrics

Recommended 0

No Suggested Reading articles found!

缓存形式	缓存	容量大小/KB	块大小/B	组数	每组块数
	L1	32	32	8	128
传统缓存	L2	64	32	16	128
	L3	128	32	32	128
	L1-draw	16	32	4	128
	L1-push	16	32	4	128
渗透缓存	L2-draw	32	32	8	128
	L2-push	32	32	8	128
	L3-draw	64	32	16	128
	L3-push	64	32	16	128

缓存形式	渗透缓存层级	访问次数	命中次数	命中率	总命中率
	L1-draw	1 000 000	645 076	64.508%
	L1-push	35 4924	46 151	13.003%
渗透缓存	L2-draw	308 773	136 667	44.261%	92.336%
	L2-push	172 106	9 654	5.609%
	L3-draw	162 452	82 403	50.725%
	L3-push	80 049	3 405	4.254%
	L1	1 000 000	722 311	72.231%
传统缓存	L2	277 689	96 518	34.757%	88.036%
	L3	181 171	61 534	33.965%

缓存形式	渗透缓存层级	访问次数	命中次数	命中率	总命中率
	L1-draw	877 441	3 200	0.365%
	L1-push	874 241	110	0.013%
渗透缓存	L2-draw	874 131	8 091	0.926%	92.160%
	L2-push	866 040	769 933	88.903%
	L3-draw	96 107	27 309	28.415%
	L3-push	68 798	3	0.004%
	L1	877 441	7 105	0.008%
传统缓存	L2	870 336	15 500	0.017%	0.076%
	L3	854 836	44 442	0.052%

缓存形式	渗透缓存层级	访问次数	命中次数	命中率	总命中率
	L1-draw	1 000 000	898 061	89.806%
	L1-push	101 939	97 863	96.002%
渗透缓存	L2-draw	4 076	2 330	57.164%	99.920%
	L2-push	1 746	251	14.376%
	L3-draw	1 495	573	38.328%
	L3-push	922	118	12.798%
	L1	1 000 000	899 589	89.9589%
传统缓存	L2	100 411	1 356	0.013%	90.118%
	L3	99 055	237	0.002%

Theoretical analysis for the data immigration between memory and processor percolation cache

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 11

References 16

Related Articles 1

Metrics

Recommended 0