USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS.

Susan Dickey, Allan Gottlieb, Richard Kenner

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

Abstract

The NYU Ultracomputer is an architecture for a large scale MIMD (Multiple Instruction stream, Multiple Data stream) shared memory parallel computer that may be viewed as a column of processors and a column of memory modules connected by a rectangular network of enhanced two by two buffered crossbars. The primary novelty of the design is the ability of the network to combine multiple requests directed at the same memory location, including a new coordination request, fetch-and-add. This permits task coordination to be achieved in a highly parallel manner. This report reviews the Ultracomputer architecture and system design and describes the VLSI enhanced buffered crossbars that are the key to the highly parallel behavior.

Original languageEnglish (US)
Title of host publicationUnknown Host Publication Title
EditorsCharles E. Leiserson
PublisherMIT Press
Pages299-316
Number of pages18
ISBN (Print)0262121131
StatePublished - 1986

Fingerprint

Data storage equipment
Systems analysis

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Dickey, S., Gottlieb, A., & Kenner, R. (1986). USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS. In C. E. Leiserson (Ed.), Unknown Host Publication Title (pp. 299-316). MIT Press.

USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS. / Dickey, Susan; Gottlieb, Allan; Kenner, Richard.

Unknown Host Publication Title. ed. / Charles E. Leiserson. MIT Press, 1986. p. 299-316.

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

Dickey, S, Gottlieb, A & Kenner, R 1986, USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS. in CE Leiserson (ed.), Unknown Host Publication Title. MIT Press, pp. 299-316.
Dickey S, Gottlieb A, Kenner R. USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS. In Leiserson CE, editor, Unknown Host Publication Title. MIT Press. 1986. p. 299-316
Dickey, Susan ; Gottlieb, Allan ; Kenner, Richard. / USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS. Unknown Host Publication Title. editor / Charles E. Leiserson. MIT Press, 1986. pp. 299-316
@inproceedings{560edec6e9ce4856b9f6717eec71582c,
title = "USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS.",
abstract = "The NYU Ultracomputer is an architecture for a large scale MIMD (Multiple Instruction stream, Multiple Data stream) shared memory parallel computer that may be viewed as a column of processors and a column of memory modules connected by a rectangular network of enhanced two by two buffered crossbars. The primary novelty of the design is the ability of the network to combine multiple requests directed at the same memory location, including a new coordination request, fetch-and-add. This permits task coordination to be achieved in a highly parallel manner. This report reviews the Ultracomputer architecture and system design and describes the VLSI enhanced buffered crossbars that are the key to the highly parallel behavior.",
author = "Susan Dickey and Allan Gottlieb and Richard Kenner",
year = "1986",
language = "English (US)",
isbn = "0262121131",
pages = "299--316",
editor = "Leiserson, {Charles E.}",
booktitle = "Unknown Host Publication Title",
publisher = "MIT Press",

}

TY - GEN

T1 - USING VLSI TO REDUCE SERIALIZATION AND MEMORY TRAFFIC IN SHARED MEMORY PARALLEL COMPUTERS.

AU - Dickey, Susan

AU - Gottlieb, Allan

AU - Kenner, Richard

PY - 1986

Y1 - 1986

N2 - The NYU Ultracomputer is an architecture for a large scale MIMD (Multiple Instruction stream, Multiple Data stream) shared memory parallel computer that may be viewed as a column of processors and a column of memory modules connected by a rectangular network of enhanced two by two buffered crossbars. The primary novelty of the design is the ability of the network to combine multiple requests directed at the same memory location, including a new coordination request, fetch-and-add. This permits task coordination to be achieved in a highly parallel manner. This report reviews the Ultracomputer architecture and system design and describes the VLSI enhanced buffered crossbars that are the key to the highly parallel behavior.

AB - The NYU Ultracomputer is an architecture for a large scale MIMD (Multiple Instruction stream, Multiple Data stream) shared memory parallel computer that may be viewed as a column of processors and a column of memory modules connected by a rectangular network of enhanced two by two buffered crossbars. The primary novelty of the design is the ability of the network to combine multiple requests directed at the same memory location, including a new coordination request, fetch-and-add. This permits task coordination to be achieved in a highly parallel manner. This report reviews the Ultracomputer architecture and system design and describes the VLSI enhanced buffered crossbars that are the key to the highly parallel behavior.

UR - http://www.scopus.com/inward/record.url?scp=0022902330&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0022902330&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0262121131

SP - 299

EP - 316

BT - Unknown Host Publication Title

A2 - Leiserson, Charles E.

PB - MIT Press

ER -