Solving the Boltzmann equation at 61 Gigaflops on a 1024-Node CM-5

L. N. Long, J. Myczkowski

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

This paper documents the use of a massively parallel computer, specifically the Connection Machine CM-5, to solve the Boltzmann equation to model one-dimensional shock wave structure, a boundary layer, and general 3-D flow fields. The Bhatnagar-Gross-Krook (BGK) model for the collision term combined with a finite difference scheme was used to model the flow. This collision terms requires accurate knoweldge of the density, temperature, and mean velocity. Great care must be taken in their calculation to insure conservation, which proved to be the most difficult part. The algorithm, however, is well suited to the Connection Machine, and accurate results were obtained with great efficiency. The 1-D version of the code (which actually models a 5-D problem in phase space) was optimized for the CM-5 and sustained 61 gigaflops on a 1024-node CM-5.

Original languageEnglish (US)
Title of host publicationProceedings of the Supercomputing Conference
PublisherPubl by IEEE
Pages528-534
Number of pages7
ISBN (Print)0818643404, 9780818643408
DOIs
StatePublished - 1993
EventProceedings of the Supercomputing '93 Conference - Portland, OR, USA
Duration: Nov 15 1993Nov 19 1993

Publication series

NameProceedings of the Supercomputing Conference
ISSN (Print)1063-9535

Other

OtherProceedings of the Supercomputing '93 Conference
CityPortland, OR, USA
Period11/15/9311/19/93

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Solving the Boltzmann equation at 61 Gigaflops on a 1024-Node CM-5'. Together they form a unique fingerprint.

  • Cite this

    Long, L. N., & Myczkowski, J. (1993). Solving the Boltzmann equation at 61 Gigaflops on a 1024-Node CM-5. In Proceedings of the Supercomputing Conference (pp. 528-534). (Proceedings of the Supercomputing Conference). Publ by IEEE. https://doi.org/10.1145/169627.169795