Hybrid partitioning for particle-in-cell simulation on shared memory systems

Author

Azari, Nasser G. ; Lee, Soo-Young

Author_Institution

Sch. of Electr. Eng., Cornell Univ., Ithaca, NY, USA

fYear

1991

fDate

20-24 May 1991

Firstpage

526

Lastpage

533

Abstract

The authors explore parallel processing issues, such as relationships between speedup and problem partitioning schemes, and problem size and time duration of each iteration for PIC (particle in cell) method on different multiprocessors. A partitioning scheme, hybrid partitioning, is introduced. Hybrid partitioning has evolved out of two general approaches to PIC problem decomposition on multiprocessors, partitioning particles and partitioning the space. The authors chose the shared memory multiprocessor environment for analyzing the parallel (distributed computing) algorithms. Two different BBN Butterfly machines (GP1000 and TC2000) were employed as testbeds

Keywords

circuit analysis computing; circuit layout CAD; parallel algorithms; BBN Butterfly machines; GP1000; TC2000; multiprocessors; parallel algorithms; particle-in-cell simulation; performance analysis; problem decomposition; problem partitioning; problem size; shared memory multiprocessor environment; shared memory systems; speedup; statistical models; time duration; Binary search trees; Computational modeling; Particle scattering; Partitioning algorithms; Physics; Plasma devices; Plasma simulation; Semiconductor devices; Supercomputers; System testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Distributed Computing Systems, 1991., 11th International Conference on

Conference_Location

Arlington, TX

Print_ISBN

0-8186-2144-3

Type

conf

DOI

10.1109/ICDCS.1991.148721

Filename

148721