### Abstract

A method of deterministic simulation of fully operational parallel machines on the analogous machines prone to errors is developed. The simulation is presented for the exclusive-read exclusive-write (EREW) PRAM and the Optical Communication Parallel Computer (OCPC), but it applies to a large class of parallel computers. It is shown that simulations of operational multiprocessor machines on faulty ones can be performed with logarithmic slowdown in the worst case. More precisely, we prove that both a PRAM with a bounded fraction of faulty processors and memory cells and an OCPC with a bounded fraction of faulty processors can simulate deterministically their fault-free counterparts with O(logn) slowdown and preprocessing done in time O(log^{2} n). The fault model is as follows. The faults are deterministic (worst-case distribution) and static (do not change in the course of a computation). If a processor attempts to communicate with some other processor (in the case of an OCPC) or read a memory word (in the case of a PRAM) then it is immediately notified whether the operation was successful (fault-free addressee) or failed (faulty addressee). This is for the first time that a general fast deterministic simulation technique is designed for the EREW PRAM with the worst-case fault distribution. The simulation is designed in such a way that it relies only on a fraction of all the operational processors. During preprocessing, the active processors retrieve the original input provided to all processors before the simulation started. This is accomplished by adapting the information-dispersal method.

Original language | English (US) |
---|---|

Title of host publication | Algorithms - ESA 1995 - 3rd Annual European Symposium, Proceedings |

Editors | Paul Spirakis |

Publisher | Springer Verlag |

Pages | 89-101 |

Number of pages | 13 |

ISBN (Print) | 3540603131, 9783540603139 |

State | Published - Jan 1 1995 |

Externally published | Yes |

Event | 3rd Annual European Symposium on Algorithms, ESA 1995 - Corfu, Greece Duration: Sep 25 1995 → Sep 27 1995 |

### Publication series

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

Volume | 979 |

ISSN (Print) | 0302-9743 |

ISSN (Electronic) | 1611-3349 |

### Conference

Conference | 3rd Annual European Symposium on Algorithms, ESA 1995 |
---|---|

Country | Greece |

City | Corfu |

Period | 9/25/95 → 9/27/95 |

### Fingerprint

### ASJC Scopus subject areas

- Theoretical Computer Science
- Computer Science(all)

### Cite this

*Algorithms - ESA 1995 - 3rd Annual European Symposium, Proceedings*(pp. 89-101). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 979). Springer Verlag.

**Fast deterministic simulation of computations on faulty parallel machines.** / Chlebus, Bogdan S.; Gąsieniec, Leszek; Pelc, Andrzej.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

*Algorithms - ESA 1995 - 3rd Annual European Symposium, Proceedings.*Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 979, Springer Verlag, pp. 89-101, 3rd Annual European Symposium on Algorithms, ESA 1995, Corfu, Greece, 9/25/95.

}

TY - GEN

T1 - Fast deterministic simulation of computations on faulty parallel machines

AU - Chlebus, Bogdan S.

AU - Gąsieniec, Leszek

AU - Pelc, Andrzej

PY - 1995/1/1

Y1 - 1995/1/1

N2 - A method of deterministic simulation of fully operational parallel machines on the analogous machines prone to errors is developed. The simulation is presented for the exclusive-read exclusive-write (EREW) PRAM and the Optical Communication Parallel Computer (OCPC), but it applies to a large class of parallel computers. It is shown that simulations of operational multiprocessor machines on faulty ones can be performed with logarithmic slowdown in the worst case. More precisely, we prove that both a PRAM with a bounded fraction of faulty processors and memory cells and an OCPC with a bounded fraction of faulty processors can simulate deterministically their fault-free counterparts with O(logn) slowdown and preprocessing done in time O(log2 n). The fault model is as follows. The faults are deterministic (worst-case distribution) and static (do not change in the course of a computation). If a processor attempts to communicate with some other processor (in the case of an OCPC) or read a memory word (in the case of a PRAM) then it is immediately notified whether the operation was successful (fault-free addressee) or failed (faulty addressee). This is for the first time that a general fast deterministic simulation technique is designed for the EREW PRAM with the worst-case fault distribution. The simulation is designed in such a way that it relies only on a fraction of all the operational processors. During preprocessing, the active processors retrieve the original input provided to all processors before the simulation started. This is accomplished by adapting the information-dispersal method.

AB - A method of deterministic simulation of fully operational parallel machines on the analogous machines prone to errors is developed. The simulation is presented for the exclusive-read exclusive-write (EREW) PRAM and the Optical Communication Parallel Computer (OCPC), but it applies to a large class of parallel computers. It is shown that simulations of operational multiprocessor machines on faulty ones can be performed with logarithmic slowdown in the worst case. More precisely, we prove that both a PRAM with a bounded fraction of faulty processors and memory cells and an OCPC with a bounded fraction of faulty processors can simulate deterministically their fault-free counterparts with O(logn) slowdown and preprocessing done in time O(log2 n). The fault model is as follows. The faults are deterministic (worst-case distribution) and static (do not change in the course of a computation). If a processor attempts to communicate with some other processor (in the case of an OCPC) or read a memory word (in the case of a PRAM) then it is immediately notified whether the operation was successful (fault-free addressee) or failed (faulty addressee). This is for the first time that a general fast deterministic simulation technique is designed for the EREW PRAM with the worst-case fault distribution. The simulation is designed in such a way that it relies only on a fraction of all the operational processors. During preprocessing, the active processors retrieve the original input provided to all processors before the simulation started. This is accomplished by adapting the information-dispersal method.

UR - http://www.scopus.com/inward/record.url?scp=84947710857&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84947710857&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84947710857

SN - 3540603131

SN - 9783540603139

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

SP - 89

EP - 101

BT - Algorithms - ESA 1995 - 3rd Annual European Symposium, Proceedings

A2 - Spirakis, Paul

PB - Springer Verlag

ER -