Data communication in hypercubes

Youcef Saad; Martin H. Schultz

doi:10.1016/0743-7315(89)90045-2

Data communication in hypercubes

Youcef Saad, Martin H. Schultz

Computer Science and Engineering

Research output: Contribution to journal › Article › peer-review

85 Scopus citations

Abstract

In this paper we consider several algorithms for exchanging data among processors in a hypercube network. The data transfer problems considered are those arising from classical numerical algorithms such as Gaussian elimination, conjugate gradient methods, and the N-body problem. We propose some estimates for the timings of the various algorithms which reveal that multiprocessors based on the hypercube topology can be very efficient in performing data exchange operations.

Original language	English (US)
Pages (from-to)	115-135
Number of pages	21
Journal	Journal of Parallel and Distributed Computing
Volume	6
Issue number	1
DOIs	https://doi.org/10.1016/0743-7315(89)90045-2
State	Published - Feb 1989

Bibliographical note

Funding Information:
* This work was supported in part by ONR Grant NO00 14-82-K-O 184 and in part by a joint study with IBM/Kingston.

Access

10.1016/0743-7315(89)90045-2

Fingerprint Dive into the research topics of 'Data communication in hypercubes'. Together they form a unique fingerprint.

Cite this

@article{2c3574d7856848a78fe32907e0db9ccc,

title = "Data communication in hypercubes",

abstract = "In this paper we consider several algorithms for exchanging data among processors in a hypercube network. The data transfer problems considered are those arising from classical numerical algorithms such as Gaussian elimination, conjugate gradient methods, and the N-body problem. We propose some estimates for the timings of the various algorithms which reveal that multiprocessors based on the hypercube topology can be very efficient in performing data exchange operations.",

author = "Youcef Saad and Schultz, {Martin H.}",

note = "Funding Information: * This work was supported in part by ONR Grant NO00 14-82-K-O 184 and in part by a joint study with IBM/Kingston.",

year = "1989",

month = feb,

doi = "10.1016/0743-7315(89)90045-2",

language = "English (US)",

volume = "6",

pages = "115--135",

journal = "Journal of Parallel and Distributed Computing",

issn = "0743-7315",

publisher = "Academic Press Inc.",

number = "1",

}

TY - JOUR

T1 - Data communication in hypercubes

AU - Saad, Youcef

AU - Schultz, Martin H.

N1 - Funding Information: * This work was supported in part by ONR Grant NO00 14-82-K-O 184 and in part by a joint study with IBM/Kingston.

PY - 1989/2

Y1 - 1989/2

N2 - In this paper we consider several algorithms for exchanging data among processors in a hypercube network. The data transfer problems considered are those arising from classical numerical algorithms such as Gaussian elimination, conjugate gradient methods, and the N-body problem. We propose some estimates for the timings of the various algorithms which reveal that multiprocessors based on the hypercube topology can be very efficient in performing data exchange operations.

AB - In this paper we consider several algorithms for exchanging data among processors in a hypercube network. The data transfer problems considered are those arising from classical numerical algorithms such as Gaussian elimination, conjugate gradient methods, and the N-body problem. We propose some estimates for the timings of the various algorithms which reveal that multiprocessors based on the hypercube topology can be very efficient in performing data exchange operations.

UR - http://www.scopus.com/inward/record.url?scp=0001755882&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0001755882&partnerID=8YFLogxK

U2 - 10.1016/0743-7315(89)90045-2

DO - 10.1016/0743-7315(89)90045-2

M3 - Article

AN - SCOPUS:0001755882

VL - 6

SP - 115

EP - 135

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 1

ER -