Load Balancing and Data Locality of the Near-Field Interaction in the Parallelization of MLFMA
This paper describes a simplified parallelization ofthe near-field interaction in the multilevel fast multipolealgorithms (MLFMA) based on a data decomposition strategywith an improved load balancing and the data locality.It showsgood performance on distributed memory computers using themessage passing interface (MPI).
MLFMA Iterative technique Parallel computing Computational electromagnetics
Wang XinGang Tang HuaNing Zhi XiaoLi Ni WeiLi Tong WeiQin
Department of Cornputer Engineer and Science,Shanghai University,Shanghai 200072,China;College of In Department of Cornputer Engineer and Science,Shanghai University,Shanghai 200072,China School of Communication and Information Engineering,Shanghai University,Shanghai 200072,China
国际会议
2008 China-Japan Joint Microwave Conference(2008年中日微波会议)
上海
英文
347-349
2008-09-10(万方平台首次上网日期,不代表论文的发表时间)