Parallelization of the multilevel fast multipole algorithm by combined use of OpenMP and VALU hardware acceleration

Jinbo Liu; Mang He; Kang Zhang; Binbin Wang; Qingwei Qiu

doi:10.1109/TAP.2014.2321141

Parallelization of the multilevel fast multipole algorithm by combined use of OpenMP and VALU hardware acceleration

Jinbo Liu, Mang He^*, Kang Zhang, Binbin Wang, Qingwei Qiu

^*Corresponding author for this work

School of Integrated Circuits and Electronics

Beijing Institute of Technology

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

Abstract

A parallel scheme that combines the OpenMP and the vector arithmetic logic unit (VALU) hardware acceleration is presented to speed up the multilevel fast multipole algorithm (MLFMA) on shared-memory computers. Performance of the hybrid parallel OpenMP-VALU MLFMA is investigated and several strategies are employed to improve the overall speedup and parallel efficiency. Effectiveness of the hybrid parallel scheme is verified by numerical results of the electromagnetic (EM) scattering examples, and related numerical stability issue is discussed as well.

Original language	English
Article number	6808505
Pages (from-to)	3884-3889
Number of pages	6
Journal	IEEE Transactions on Antennas and Propagation
Volume	62
Issue number	7
DOIs	https://doi.org/10.1109/TAP.2014.2321141
Publication status	Published - Jul 2014

Keywords

Electromagnetic scattering
OpenMP
multilevel fast multipole algorithm
parallel computing
vector arithmetic logic unit (VALU)

Access to Document

10.1109/TAP.2014.2321141

Cite this

@article{ad82cc07bff24633a0a5346aca246048,

title = "Parallelization of the multilevel fast multipole algorithm by combined use of OpenMP and VALU hardware acceleration",

abstract = "A parallel scheme that combines the OpenMP and the vector arithmetic logic unit (VALU) hardware acceleration is presented to speed up the multilevel fast multipole algorithm (MLFMA) on shared-memory computers. Performance of the hybrid parallel OpenMP-VALU MLFMA is investigated and several strategies are employed to improve the overall speedup and parallel efficiency. Effectiveness of the hybrid parallel scheme is verified by numerical results of the electromagnetic (EM) scattering examples, and related numerical stability issue is discussed as well.",

keywords = "Electromagnetic scattering, OpenMP, multilevel fast multipole algorithm, parallel computing, vector arithmetic logic unit (VALU)",

author = "Jinbo Liu and Mang He and Kang Zhang and Binbin Wang and Qingwei Qiu",

year = "2014",

month = jul,

doi = "10.1109/TAP.2014.2321141",

language = "English",

volume = "62",

pages = "3884--3889",

journal = "IEEE Transactions on Antennas and Propagation",

issn = "0018-926X",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "7",

}

TY - JOUR

T1 - Parallelization of the multilevel fast multipole algorithm by combined use of OpenMP and VALU hardware acceleration

AU - Liu, Jinbo

AU - He, Mang

AU - Zhang, Kang

AU - Wang, Binbin

AU - Qiu, Qingwei

PY - 2014/7

Y1 - 2014/7

N2 - A parallel scheme that combines the OpenMP and the vector arithmetic logic unit (VALU) hardware acceleration is presented to speed up the multilevel fast multipole algorithm (MLFMA) on shared-memory computers. Performance of the hybrid parallel OpenMP-VALU MLFMA is investigated and several strategies are employed to improve the overall speedup and parallel efficiency. Effectiveness of the hybrid parallel scheme is verified by numerical results of the electromagnetic (EM) scattering examples, and related numerical stability issue is discussed as well.

AB - A parallel scheme that combines the OpenMP and the vector arithmetic logic unit (VALU) hardware acceleration is presented to speed up the multilevel fast multipole algorithm (MLFMA) on shared-memory computers. Performance of the hybrid parallel OpenMP-VALU MLFMA is investigated and several strategies are employed to improve the overall speedup and parallel efficiency. Effectiveness of the hybrid parallel scheme is verified by numerical results of the electromagnetic (EM) scattering examples, and related numerical stability issue is discussed as well.

KW - Electromagnetic scattering

KW - OpenMP

KW - multilevel fast multipole algorithm

KW - parallel computing

KW - vector arithmetic logic unit (VALU)

UR - http://www.scopus.com/inward/record.url?scp=84904178439&partnerID=8YFLogxK

U2 - 10.1109/TAP.2014.2321141

DO - 10.1109/TAP.2014.2321141

M3 - Article

AN - SCOPUS:84904178439

SN - 0018-926X

VL - 62

SP - 3884

EP - 3889

JO - IEEE Transactions on Antennas and Propagation

JF - IEEE Transactions on Antennas and Propagation

IS - 7

M1 - 6808505

ER -

Parallelization of the multilevel fast multipole algorithm by combined use of OpenMP and VALU hardware acceleration

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this