Generalized schemes for access and alignment of data in parallel processors with self-routing interconnection networks
Journal Article
·
· Journal of Parallel and Distributed Computing; (United States)
- Dept. of Electrical Engineering-Systems, Univ. of Southern California, Los Angeles, CA (US)
In this paper the authors give a generalized solution to the problem of conflict-free access of various templates of data of a matrix, when they are stored in memory units in a parallel processor. The important features of the method are: compact representation of a skewing scheme, simple address computation, use of self- routing schemes to set up the interconnection network, and a general framework for the study of skewing schemes. In the method, each template access of interest will be a linear permutation on the processor address. The linear permutation involved determines the types of templates accessible. For parallel access of the most important templates, namely, row, column, main diagonal, and square blocks, the interconnection network needs to realize only the class of linear-complement permutations. It is known that with Benes or Omega as the interconnection network, one can efficiently self-route these permutations; this compares favorably with the schemes proposed by other researchers who assume that a cross bar is available for processor-memory interconnections. Hence, the approach given in the paper can be used to solve the data alignment problem for the existing parallel machines such as IBM RP3, Cedar multiprocessor, and NYU Ultracomputer. This is a generalized solution to the data skewing problem and encompasses the previous efforts by other researchers as special cases.
- Sponsoring Organization:
- NSF; National Science Foundation, Washington, DC (United States)
- OSTI ID:
- 5071023
- Journal Information:
- Journal of Parallel and Distributed Computing; (United States), Journal Name: Journal of Parallel and Distributed Computing; (United States) Vol. 11:2; ISSN 0743-7315; ISSN JPDCE
- Country of Publication:
- United States
- Language:
- English
Similar Records
1989 International Conference on Parallel Processing, University Park, PA, Aug. 8-12, 1989, Proceedings. Volume 1 - Architecture
Interconnection networks: improved routing methods for Clos and Benes networks
Design of an array processor for image processing
Conference
·
Sat Dec 31 23:00:00 EST 1988
·
OSTI ID:6686984
Interconnection networks: improved routing methods for Clos and Benes networks
Thesis/Dissertation
·
Wed Dec 31 23:00:00 EST 1986
·
OSTI ID:6445478
Design of an array processor for image processing
Journal Article
·
Thu Jan 31 23:00:00 EST 1991
· Journal of Parallel and Distributed Computing; (United States)
·
OSTI ID:5001355