HPL performance prevision to intending system improvement

CORC > 计算技术研究所 > 中国科学院计算技术研究所 > 中国科学院计算技术研究所期刊论文 > 英文

	HPL performance prevision to intending system improvement
	Zhang, WL ; Chen, MY ; Fan, JP
刊名	PARALLEL AND DISTRIBUTED PROCESSING AND APPLICATIONS, PROCEEDINGS
	2004
卷号	3358 页码:777-782
ISSN号	0302-9743
英文摘要	HPL is a parallel Unpack benchmark package widely adopted in massive cluster system performance test. On HPL data layout among processors, a law to determine block size NB theoretically, which breaks through dependence on trial-and-error experiments, is found based on in-depth analysis of blocked parallel solution algorithm of linear algebra equations and implementation mechanics in HPL. According to that law. an emulation model to toughly estimate HPL execution time is constructed. Verified by real system. the model is used to do some scientific prevision on the benefits to Linpack test brought by intending system improvement, such as respectively memory size increase, communication bandwidth increase and so on. It is expected to conduce to direct system improvement on optimizing HPL test in the future.
WOS研究方向	Computer Science
语种	英语
出版者	SPRINGER-VERLAG BERLIN
WOS记录号	WOS:000226159800087
内容类型	期刊论文
源URL	[http://119.78.100.204/handle/2XEOYT63/13988]
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Zhang, WL
作者单位	1.Chinese Acad Sci, Inst Comp Technol, Natl Res Ctr Intelligent Comp Syst, NCIC, Beijing 100080, Peoples R China 2.Chinese Acad Sci, Grad Sch, NCIC, Beijing 100080, Peoples R China
推荐引用方式 GB/T 7714	Zhang, WL,Chen, MY,Fan, JP. HPL performance prevision to intending system improvement[J]. PARALLEL AND DISTRIBUTED PROCESSING AND APPLICATIONS, PROCEEDINGS,2004,3358:777-782.
APA	Zhang, WL,Chen, MY,&Fan, JP.(2004).HPL performance prevision to intending system improvement.PARALLEL AND DISTRIBUTED PROCESSING AND APPLICATIONS, PROCEEDINGS,3358,777-782.
MLA	Zhang, WL,et al."HPL performance prevision to intending system improvement".PARALLEL AND DISTRIBUTED PROCESSING AND APPLICATIONS, PROCEEDINGS 3358(2004):777-782.