CFD on parallel computers

Basu, AJ (1994) CFD on parallel computers. In: CFD: Advances and Applications, India.

Full text not available from this repository.

Abstract

CFD or Computational Fluid Dynamics is one of the scientific disciplines that has always posed new challenges to the capabilities of the modern, ultra-fast supercomputers, and now to the even faster parallel computers. For applications where number crunching is of primary importance, there is perhaps no escaping parallel computers since sequential computers can only be (as projected) as fast as a few gigaflops and no more, unless, of course, some altogether new technology appears in future. For parallel computers, on the other hand, there is no such limit since any number of processors can be made to work in parallel. Computationally demanding CFD codes and parallel computers are therefore soul-mates, and will remain so for all foreseeable future. So much so that there is a separate and fast-emerging discipline that tackles problems specific to CFD as applied to parallel computers. For some years now, there is an international conference on parallel CFD. So, one can indeed say that parallel CFD has arrived. To understand how CFD codes are parallelized, one must understand a little about how parallel computers function. Therefore, in what follows we will first deal with parallel computers, how a typical CFD code (if there is one such) looks like, and then the strategies of parallelization.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Computational fluid dynamics;Parallel processing
Subjects: ENGINEERING > Fluid Mechanics and Thermodynamics
Division/Department: Computational and Theoretical Fluid Dynamics Division
Depositing User: M/S ICAST NAL
Date Deposited: 03 Oct 2006
Last Modified: 24 May 2010 09:50
URI: http://nal-ir.nal.res.in/id/eprint/2819

Actions (login required)

View Item