Parallel implementation of w-projection wide-field imaging

doi:10.1016/j.scib.2019.04.004

	Parallel implementation of w-projection wide-field imaging
	Lao, Baoqiang 1; An, Tao 1; Yu, Ang 2; Zhang, Wenhui 2,3; Wang, Junyi 2; Guo, Quan 1; Guo, Shaoguang 1; Wu, Xiaocong 1
刊名	SCIENCE BULLETIN
	2019-05-15
卷号	64 期号:9 页码:586-594
关键词	Radio synthesis arrays Square kilometre array Wide field imaging Parallelization w-Projection
ISSN号	2095-9273
DOI	10.1016/j.scib.2019.04.004
通讯作者	An, Tao(antao@shao.ac.cn)
英文摘要	w-Projection is a wide-field imaging technique that is widely used in radio synthesis arrays. Processing the wide-field big data generated by the future Square Kilometre Array (SKA) will require significant updates to current methods to significantly reduce the time consumed on data processing. Data loading and gridding are found to be two major time-consuming tasks in w-projection. In this paper, we investigate two parallel methods of accelerating w-projection processing on multiple nodes: the hybrid Message Passing Interface (MPI) and Open Multi-Processing (OpenMP) method based on multicore Central Processing Units (CPUs) and the hybrid MPI and Compute Unified Device Architecture (CUDA) method based on Graphics Processing Units (GPUs). Both methods are successfully employed and operated in various computational environments, confirming their robustness. The experimental results show that the total runtime of both MPI + OpenMP and MPI + CUDA methods is significantly shorter than that of single-thread processing. MPI + CUDA generally shows faster performance when running on multiple nodes than MPI + OpenMP, especially on large numbers of nodes. The single-precision GPU-based processing yields faster computation than the double-precision processing; while the single- and doubleprecision CPU-based processing shows consistent computational performance. The gridding time remarkably increases when the support size of the convolution kernel is larger than 8 and the image size is larger than 2,048 pixels. The present research offers useful guidance for developing SKA imaging pipelines. (C) 2019 Science China Press. Published by Elsevier B.V. and Science China Press. All rights reserved.
WOS关键词	TABLE DATA SYSTEM
WOS研究方向	Science & Technology - Other Topics
语种	英语
出版者	ELSEVIER SCIENCE BV
WOS记录号	WOS:000472944600007
内容类型	期刊论文
源URL	[http://libir.pmo.ac.cn/handle/332002/27578]
专题	中国科学院紫金山天文台
通讯作者	An, Tao
作者单位	1.Chinese Acad Sci, Shanghai Astron Observ, Key Lab Radio Astron, Shanghai 200030, Peoples R China 2.Guilin Univ Elect Technol, Guangxi Cooperat Innovat Ctr Cloud Comp & Big Dat, Guilin 541004, Peoples R China 3.Guilin Univ Elect Technol, Guangxi Coll & Univ Key Lab Cloud Comp & Complex, Guilin 541004, Peoples R China
推荐引用方式 GB/T 7714	Lao, Baoqiang,An, Tao,Yu, Ang,et al. Parallel implementation of w-projection wide-field imaging[J]. SCIENCE BULLETIN,2019,64(9):586-594.
APA	Lao, Baoqiang.,An, Tao.,Yu, Ang.,Zhang, Wenhui.,Wang, Junyi.,...&Wu, Xiaocong.(2019).Parallel implementation of w-projection wide-field imaging.SCIENCE BULLETIN,64(9),586-594.
MLA	Lao, Baoqiang,et al."Parallel implementation of w-projection wide-field imaging".SCIENCE BULLETIN 64.9(2019):586-594.