Efficient parallelization of multilevel fast multipole algorithm for electromagnetic simulation on many-core SW26010 processor

Wei Jia He; Ming Lin Yang; Wu Wang; Xin Qing Sheng

doi:10.1007/s11227-020-03308-9

Efficient parallelization of multilevel fast multipole algorithm for electromagnetic simulation on many-core SW26010 processor

Wei Jia He, Ming Lin Yang^*, Wu Wang, Xin Qing Sheng

^*Corresponding author for this work

School of Integrated Circuits and Electronics

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

3 Citations (Scopus)

Abstract

A many-core parallel approach of the multilevel fast multipole algorithm (MLFMA) based on the Athread parallel programming model is presented on the homegrown many-core SW26010 CPU of China. In the proposed many-core implementation of MLFMA, the data access efficiency is improved by using data structures based on the structure of array. The adaptive workload distribution strategies are adopted on different MLFMA tree levels to ensure full utilization of computing capability and the scratchpad memory. A double buffering scheme is specially designed to make communication overlapped computation. The resulting Athread-based many-core implementation of the MLFMA is capable of solving real-life problems with over one million unknowns with a remarkable speedup. The capability and efficiency of the proposed method are analyzed through the examples of computing scattering by spheres and a practical aerocraft. Numerical results show that with the proposed parallel scheme, the total speedup ratios from 6.4 to 8.0 can be achieved, compared with the CPU master core.

Original language	English
Pages (from-to)	1502-1516
Number of pages	15
Journal	Journal of Supercomputing
Volume	77
Issue number	2
DOIs	https://doi.org/10.1007/s11227-020-03308-9
Publication status	Published - Feb 2021

Keywords

3D scattering
Many-core parallelization
Multilevel fast multipole algorithm
Surface integral equations
Sw26010 processor

Access to Document

10.1007/s11227-020-03308-9

Cite this

@article{e61b6bdd54524427bbf5a3dddc6e1fea,

title = "Efficient parallelization of multilevel fast multipole algorithm for electromagnetic simulation on many-core SW26010 processor",

abstract = "A many-core parallel approach of the multilevel fast multipole algorithm (MLFMA) based on the Athread parallel programming model is presented on the homegrown many-core SW26010 CPU of China. In the proposed many-core implementation of MLFMA, the data access efficiency is improved by using data structures based on the structure of array. The adaptive workload distribution strategies are adopted on different MLFMA tree levels to ensure full utilization of computing capability and the scratchpad memory. A double buffering scheme is specially designed to make communication overlapped computation. The resulting Athread-based many-core implementation of the MLFMA is capable of solving real-life problems with over one million unknowns with a remarkable speedup. The capability and efficiency of the proposed method are analyzed through the examples of computing scattering by spheres and a practical aerocraft. Numerical results show that with the proposed parallel scheme, the total speedup ratios from 6.4 to 8.0 can be achieved, compared with the CPU master core.",

keywords = "3D scattering, Many-core parallelization, Multilevel fast multipole algorithm, Surface integral equations, Sw26010 processor",

author = "He, {Wei Jia} and Yang, {Ming Lin} and Wu Wang and Sheng, {Xin Qing}",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Science+Business Media, LLC, part of Springer Nature.",

year = "2021",

month = feb,

doi = "10.1007/s11227-020-03308-9",

language = "English",

volume = "77",

pages = "1502--1516",

journal = "Journal of Supercomputing",

issn = "0920-8542",

publisher = "Springer Netherlands",

number = "2",

}

TY - JOUR

T1 - Efficient parallelization of multilevel fast multipole algorithm for electromagnetic simulation on many-core SW26010 processor

AU - He, Wei Jia

AU - Yang, Ming Lin

AU - Wang, Wu

AU - Sheng, Xin Qing

PY - 2021/2

Y1 - 2021/2

N2 - A many-core parallel approach of the multilevel fast multipole algorithm (MLFMA) based on the Athread parallel programming model is presented on the homegrown many-core SW26010 CPU of China. In the proposed many-core implementation of MLFMA, the data access efficiency is improved by using data structures based on the structure of array. The adaptive workload distribution strategies are adopted on different MLFMA tree levels to ensure full utilization of computing capability and the scratchpad memory. A double buffering scheme is specially designed to make communication overlapped computation. The resulting Athread-based many-core implementation of the MLFMA is capable of solving real-life problems with over one million unknowns with a remarkable speedup. The capability and efficiency of the proposed method are analyzed through the examples of computing scattering by spheres and a practical aerocraft. Numerical results show that with the proposed parallel scheme, the total speedup ratios from 6.4 to 8.0 can be achieved, compared with the CPU master core.

AB - A many-core parallel approach of the multilevel fast multipole algorithm (MLFMA) based on the Athread parallel programming model is presented on the homegrown many-core SW26010 CPU of China. In the proposed many-core implementation of MLFMA, the data access efficiency is improved by using data structures based on the structure of array. The adaptive workload distribution strategies are adopted on different MLFMA tree levels to ensure full utilization of computing capability and the scratchpad memory. A double buffering scheme is specially designed to make communication overlapped computation. The resulting Athread-based many-core implementation of the MLFMA is capable of solving real-life problems with over one million unknowns with a remarkable speedup. The capability and efficiency of the proposed method are analyzed through the examples of computing scattering by spheres and a practical aerocraft. Numerical results show that with the proposed parallel scheme, the total speedup ratios from 6.4 to 8.0 can be achieved, compared with the CPU master core.

KW - 3D scattering

KW - Many-core parallelization

KW - Multilevel fast multipole algorithm

KW - Surface integral equations

KW - Sw26010 processor

UR - http://www.scopus.com/inward/record.url?scp=85085282404&partnerID=8YFLogxK

U2 - 10.1007/s11227-020-03308-9

DO - 10.1007/s11227-020-03308-9

M3 - Article

AN - SCOPUS:85085282404

SN - 0920-8542

VL - 77

SP - 1502

EP - 1516

JO - Journal of Supercomputing

JF - Journal of Supercomputing

IS - 2

ER -

Efficient parallelization of multilevel fast multipole algorithm for electromagnetic simulation on many-core SW26010 processor

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this