Abstract
An existing computational fluid dynamics (CFD) computer code was converted from a single process execution mode to one employing both massively parallel processing (MPP) on an nCUBE MPP computer using the native nCUBE message passing application programmers interface (API) and distributed processing (DP) on a cluster of workstations using the Parallel Virtual Machine (PVM) API for message passing. Specifications for the conversion included the ability to easily incorporate new message passing API's, automatic decomposition and distribution of a problem to the different processors, direct node to node communication, and the same user interface as in the single process mode. There was also a strong desire to have a single code which could be compiled to run in any programmed message passing API, in the multitasking parallel processing mode already in the code, and in single process mode. These criteria produced a single code which eliminates the concern associated with maintaining multiple versions of a code, provides efficient communication, and minimizes the learning curve for users not conversant with the mechanics of distributed processing.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
