Solving problems with over one billion unknowns by the MLFMA

Xiao Min Pan; Wei Chao Pi; Ming Lin Yang; Zhen Peng; Xin Qing Sheng

doi:10.1109/TAP.2012.2189746

Solving problems with over one billion unknowns by the MLFMA

Xiao Min Pan^*, Wei Chao Pi, Ming Lin Yang, Zhen Peng, Xin Qing Sheng

^*Corresponding author for this work

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

91 Citations (Scopus)

Abstract

Using OpenMP to further accelerate the pure MPI parallel MLFMA, an efficient and flexible parallel multilevel fast multipole algorithm (MPI-OpenMP-MLFMA) is proposed. Compared with previous MPI parallel schemes, the MPI-OpenMP-MLFMA improves the load-balance and scalability greatly. The computational capability of the proposed MPI-OpenMP-MLFMA is demonstrated by computing scattering from two extremely large targets: a sphere with a diameter of 1200 wavelengths, modeled by 1,063,706,700 unknowns, and an airplane model with the largest dimension of 1600 wavelengths, involving 288,151,344 unknowns.

Original language	English
Article number	6164224
Pages (from-to)	2571-2574
Number of pages	4
Journal	IEEE Transactions on Antennas and Propagation
Volume	60
Issue number	5
DOIs	https://doi.org/10.1109/TAP.2012.2189746
Publication status	Published - 2012

Keywords

Electromagnetic scattering
MPI
OpenMP
parallelization
shared memory systems

Access to Document

10.1109/TAP.2012.2189746

Cite this

@article{7f9ec442f5f24707876f05dea6d53e4e,

title = "Solving problems with over one billion unknowns by the MLFMA",

abstract = "Using OpenMP to further accelerate the pure MPI parallel MLFMA, an efficient and flexible parallel multilevel fast multipole algorithm (MPI-OpenMP-MLFMA) is proposed. Compared with previous MPI parallel schemes, the MPI-OpenMP-MLFMA improves the load-balance and scalability greatly. The computational capability of the proposed MPI-OpenMP-MLFMA is demonstrated by computing scattering from two extremely large targets: a sphere with a diameter of 1200 wavelengths, modeled by 1,063,706,700 unknowns, and an airplane model with the largest dimension of 1600 wavelengths, involving 288,151,344 unknowns.",

keywords = "Electromagnetic scattering, MPI, OpenMP, parallelization, shared memory systems",

author = "Pan, {Xiao Min} and Pi, {Wei Chao} and Yang, {Ming Lin} and Zhen Peng and Sheng, {Xin Qing}",

year = "2012",

doi = "10.1109/TAP.2012.2189746",

language = "English",

volume = "60",

pages = "2571--2574",

journal = "IEEE Transactions on Antennas and Propagation",

issn = "0018-926X",

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

number = "5",

}

TY - JOUR

T1 - Solving problems with over one billion unknowns by the MLFMA

AU - Pan, Xiao Min

AU - Pi, Wei Chao

AU - Yang, Ming Lin

AU - Peng, Zhen

AU - Sheng, Xin Qing

PY - 2012

Y1 - 2012

N2 - Using OpenMP to further accelerate the pure MPI parallel MLFMA, an efficient and flexible parallel multilevel fast multipole algorithm (MPI-OpenMP-MLFMA) is proposed. Compared with previous MPI parallel schemes, the MPI-OpenMP-MLFMA improves the load-balance and scalability greatly. The computational capability of the proposed MPI-OpenMP-MLFMA is demonstrated by computing scattering from two extremely large targets: a sphere with a diameter of 1200 wavelengths, modeled by 1,063,706,700 unknowns, and an airplane model with the largest dimension of 1600 wavelengths, involving 288,151,344 unknowns.

AB - Using OpenMP to further accelerate the pure MPI parallel MLFMA, an efficient and flexible parallel multilevel fast multipole algorithm (MPI-OpenMP-MLFMA) is proposed. Compared with previous MPI parallel schemes, the MPI-OpenMP-MLFMA improves the load-balance and scalability greatly. The computational capability of the proposed MPI-OpenMP-MLFMA is demonstrated by computing scattering from two extremely large targets: a sphere with a diameter of 1200 wavelengths, modeled by 1,063,706,700 unknowns, and an airplane model with the largest dimension of 1600 wavelengths, involving 288,151,344 unknowns.

KW - Electromagnetic scattering

KW - MPI

KW - OpenMP

KW - parallelization

KW - shared memory systems

UR - http://www.scopus.com/inward/record.url?scp=84860676125&partnerID=8YFLogxK

U2 - 10.1109/TAP.2012.2189746

DO - 10.1109/TAP.2012.2189746

M3 - Article

AN - SCOPUS:84860676125

SN - 0018-926X

VL - 60

SP - 2571

EP - 2574

JO - IEEE Transactions on Antennas and Propagation

JF - IEEE Transactions on Antennas and Propagation

IS - 5

M1 - 6164224

ER -

Solving problems with over one billion unknowns by the MLFMA

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this