Solving MoM problems with million level unknowns using a parallel out-of-core solver on a high performance cluster

Author

Yu Zhang ; Sarkar, T.K. ; Taylor, M. ; Moon, H.

Author_Institution

EECS, Syracuse Univ., Syracuse, NY, USA

fYear

2009

fDate

1-5 June 2009

Firstpage

1

Lastpage

4

Abstract

A newly developed code, TIDES, is run on a high performance computing cluster for complicated EM problems. TIDES has the capability to exceed the memory limitations of typical computer systems by using the method of moments with higher-order basis functions and a parallel out-of-core solver. In 2008, TIDES solved an MoM problem with half a million unknowns using only 320 GB RAM and 4 TB of hard disk space. A project with million level unknowns is simulated in this paper to create a new record by using only 1 TB RAM with approximately 16 TB RAM of hard disk space.

Keywords

computational electromagnetics; hard discs; integral equations; method of moments; parallel processing; random-access storage; MoM; RAM; TIDES code; complicated EM problem; hard disk space; high performance computing cluster; higher-order basis function; integral equation; method of moment; million level unknown; parallel out-of-core solver; Antenna arrays; Concurrent computing; Hard disks; High performance computing; Moment methods; Random access memory; Read-write memory; Space technology; Synthetic aperture sonar; Tides; High performance computing; Method of Moments; Parallel Out-of-Core Solver;

fLanguage

English

Publisher

ieee

Conference_Titel

Antennas and Propagation Society International Symposium, 2009. APSURSI '09. IEEE

Conference_Location

Charleston, SC

ISSN

1522-3965

Print_ISBN

978-1-4244-3647-7

Type

conf

DOI

10.1109/APS.2009.5171723

Filename

5171723