Shared-memory simulations on a faulty-memory DMM

Bogdan S. Chlebus, Anna Gambin, Piotr Indyk

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

Abstract

We study the Distributed Memory Machine (DMM) with faults in memory. The DMM consists of n synchronized processors together with n memory units (MUs). A MU can be accessed by at most one processor at a time. The total number of memory faults is assumed to be at most a fixed fraction of the total number of words. We develop two fast randomized simulations of the PRAM on such a faulty DMM. A simulation consists of two phases: the preprocessing is followed by the simulation proper done in a step-by-step fashion. One simulation is of an n log n-processor PRAM and it operates with the optimal expected slowdown O(log n), the other is of a PRAM with n/log n processors and has the slowdown O(log log n).

Original languageEnglish (US)
Title of host publicationAutomata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings
EditorsFriedhelm Meyer auf der Heide, Burkhard Monien
PublisherSpringer Verlag
Pages586-597
Number of pages12
ISBN (Print)3540614400, 9783540614401
StatePublished - Jan 1 1996
Externally publishedYes
Event23rd International Colloquium on Automata, Languages, and Programming, ICALP 1996 - Paderborn, Germany
Duration: Jul 8 1996Jul 12 1996

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1099
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference23rd International Colloquium on Automata, Languages, and Programming, ICALP 1996
CountryGermany
CityPaderborn
Period7/8/967/12/96

Fingerprint

Distributed Memory
Shared Memory
Data storage equipment
Simulation
Fault
Unit
Preprocessing

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Chlebus, B. S., Gambin, A., & Indyk, P. (1996). Shared-memory simulations on a faulty-memory DMM. In F. Meyer auf der Heide, & B. Monien (Eds.), Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings (pp. 586-597). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1099). Springer Verlag.

Shared-memory simulations on a faulty-memory DMM. / Chlebus, Bogdan S.; Gambin, Anna; Indyk, Piotr.

Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings. ed. / Friedhelm Meyer auf der Heide; Burkhard Monien. Springer Verlag, 1996. p. 586-597 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1099).

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

Chlebus, BS, Gambin, A & Indyk, P 1996, Shared-memory simulations on a faulty-memory DMM. in F Meyer auf der Heide & B Monien (eds), Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1099, Springer Verlag, pp. 586-597, 23rd International Colloquium on Automata, Languages, and Programming, ICALP 1996, Paderborn, Germany, 7/8/96.
Chlebus BS, Gambin A, Indyk P. Shared-memory simulations on a faulty-memory DMM. In Meyer auf der Heide F, Monien B, editors, Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings. Springer Verlag. 1996. p. 586-597. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Chlebus, Bogdan S. ; Gambin, Anna ; Indyk, Piotr. / Shared-memory simulations on a faulty-memory DMM. Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings. editor / Friedhelm Meyer auf der Heide ; Burkhard Monien. Springer Verlag, 1996. pp. 586-597 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{f3c812fea1b946728c56ba440bbd4c9e,
title = "Shared-memory simulations on a faulty-memory DMM",
abstract = "We study the Distributed Memory Machine (DMM) with faults in memory. The DMM consists of n synchronized processors together with n memory units (MUs). A MU can be accessed by at most one processor at a time. The total number of memory faults is assumed to be at most a fixed fraction of the total number of words. We develop two fast randomized simulations of the PRAM on such a faulty DMM. A simulation consists of two phases: the preprocessing is followed by the simulation proper done in a step-by-step fashion. One simulation is of an n log n-processor PRAM and it operates with the optimal expected slowdown O(log n), the other is of a PRAM with n/log n processors and has the slowdown O(log log n).",
author = "Chlebus, {Bogdan S.} and Anna Gambin and Piotr Indyk",
year = "1996",
month = "1",
day = "1",
language = "English (US)",
isbn = "3540614400",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "586--597",
editor = "{Meyer auf der Heide}, Friedhelm and Burkhard Monien",
booktitle = "Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings",

}

TY - GEN

T1 - Shared-memory simulations on a faulty-memory DMM

AU - Chlebus, Bogdan S.

AU - Gambin, Anna

AU - Indyk, Piotr

PY - 1996/1/1

Y1 - 1996/1/1

N2 - We study the Distributed Memory Machine (DMM) with faults in memory. The DMM consists of n synchronized processors together with n memory units (MUs). A MU can be accessed by at most one processor at a time. The total number of memory faults is assumed to be at most a fixed fraction of the total number of words. We develop two fast randomized simulations of the PRAM on such a faulty DMM. A simulation consists of two phases: the preprocessing is followed by the simulation proper done in a step-by-step fashion. One simulation is of an n log n-processor PRAM and it operates with the optimal expected slowdown O(log n), the other is of a PRAM with n/log n processors and has the slowdown O(log log n).

AB - We study the Distributed Memory Machine (DMM) with faults in memory. The DMM consists of n synchronized processors together with n memory units (MUs). A MU can be accessed by at most one processor at a time. The total number of memory faults is assumed to be at most a fixed fraction of the total number of words. We develop two fast randomized simulations of the PRAM on such a faulty DMM. A simulation consists of two phases: the preprocessing is followed by the simulation proper done in a step-by-step fashion. One simulation is of an n log n-processor PRAM and it operates with the optimal expected slowdown O(log n), the other is of a PRAM with n/log n processors and has the slowdown O(log log n).

UR - http://www.scopus.com/inward/record.url?scp=72849110811&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=72849110811&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:72849110811

SN - 3540614400

SN - 9783540614401

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 586

EP - 597

BT - Automata, Languages and Programming - 23rd International Colloquium, ICALP 1996, Proceedings

A2 - Meyer auf der Heide, Friedhelm

A2 - Monien, Burkhard

PB - Springer Verlag

ER -