YNAO OpenIR  > 其他
一种基于MPI+CUDA的高性能天文数据分发方法
Alternative TitleA High-Performance Distribution Method of Astronomical Data Based on MPI +CUDA
方兵1; 邓辉1; 张晓丽1; 梅盈2; 石聪明1; 陈晓林1; 戴伟1,2; 吴静平1; 王锋1,2
Source Publication天文研究与技术(Astronomical Research & Technology)
2017-10
Volume14Issue:4Pages:481-487
DOI10.14005/j.cnki.issn1672-7673.20170921.001
ClassificationTP311.1
Contribution Rank第2完成单位
Indexed ByCSCD
Keyword图像重建 mpi+gpu 数据分发
Abstract

天文海量数据的出现给天文软件的开发带来了诸多挑战.近年来,随着并行计算技术的发展,MPI+GPU逐渐成为当前高性能天文数据处理的主要模式.针对太阳高分辨图像重建中如何提高重建性能这一问题,对其中的数据读取与数据分发方法进行了系统研究.传统的MPI并行处理时,主进程将原图切割成子块,随后将子块发送到各子进程重建,重建后的结果返回主进程.当子进程数量较大且计算节点数量较少时,这种数据分发过程显著增加通讯时长,影响整个重建过程的效率.提出MPI+CUDA的一种树状数据分发方法,给出了算法的基本思路与实现方法.实验结果表明,树状分发方式比一般采用的平行分发方式可以提高速度近一倍,成果对天文海量数据开发处理有一定的借鉴作用. 

Other Abstract

The appearance of massive astronomical data has brought a lot of challenges to the development of astronomy software. In recent years, with the development of parallel computing technology, MPI+GPU mode has become the main mode for current high performance astronomical data processing gradually. For the problem of how to improve reconstruction performance in reconstruction of solar high resolution image, this paper has made systematic research on data reading and data distribution method. During traditional MPI parallel processing, master process cuts original image into sub-blocks, and then delivers sub-blocks into each sub-process for reconstruction, the results after reconstruction will be returned to the master process. When the number of sub-process is big and calculation nodes are few, this data distribution process will increase the time for communication significantly and affect the efficiency of the whole reconstruction process. This paper has proposed a tree data distribution method under MPI+CUDA and offered basic ideas and realization method of the algorithm. Experimental results have shown that tree distribution mode nearly doubles the speed than generally adopted plane distribution mode; its achievements provide certain reference for the development and processing of massive astronomical data.

Funding Project国家自然科学基金[U1531132] ; 国家自然科学基金[U1631129] ; 国家自然科学基金[11403009]
Funding Organization国家自然科学基金[U1531132, U1631129, 11403009]
Language中文
Subject Area天文学 ; 天文学其他学科 ; 计算机科学技术 ; 计算机应用
MOST Discipline Catalogue理学 ; 理学::天文学
ISSN1672-7673
CSCD IDCSCD:6071977
Citation statistics
Document Type期刊论文
Identifierhttp://ir.ynao.ac.cn/handle/114a53/10160
Collection其他
Affiliation1.昆明理工大学云南省计算机技术应用重点实验室,云南,昆明,650500
2.中国科学院云南天文台,云南,昆明,650011
Recommended Citation
GB/T 7714
方兵,邓辉,张晓丽,等. 一种基于MPI+CUDA的高性能天文数据分发方法[J]. 天文研究与技术(Astronomical Research & Technology),2017,14(4):481-487.
APA 方兵.,邓辉.,张晓丽.,梅盈.,石聪明.,...&王锋.(2017).一种基于MPI+CUDA的高性能天文数据分发方法.天文研究与技术(Astronomical Research & Technology),14(4),481-487.
MLA 方兵,et al."一种基于MPI+CUDA的高性能天文数据分发方法".天文研究与技术(Astronomical Research & Technology) 14.4(2017):481-487.
Files in This Item:
File Name/Size DocType Version Access License
一种基于MPI CUDA的高性能天文数据(393KB)期刊论文出版稿开放获取CC BY-NC-SAView Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[方兵]'s Articles
[邓辉]'s Articles
[张晓丽]'s Articles
Baidu academic
Similar articles in Baidu academic
[方兵]'s Articles
[邓辉]'s Articles
[张晓丽]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[方兵]'s Articles
[邓辉]'s Articles
[张晓丽]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 一种基于MPI CUDA的高性能天文数据分发方法.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.