Optimal Algorithms for Synthesis of Reliable Application-Specific Heterogeneous Multiprocessors

Aurobindo Dasgupta, Ramesh Karri

Research output: Contribution to journalArticle

Abstract

Fast and optimally-reliable application-specific multiprocessor-synthesis is critical in system-level design, especially in medical, automotive, space, and military applications. Previous work in multiprocessor-synthesis & task-allocation for performance & reliability requires exponential time, and therefore, is useful only for small examples. We present the first deterministic and provably-optimal algorithm (RELSYN-OPT) to synthesize real-time, reliable multiprocessors using a heterogeneous library of N processors and L link types. We prove that for a series-parallel graph with M subtasks and nested-depth d, the worst-case computational complexity of RELSYN-OPT is O(M·(L+N)·Nd). For tree-structured task graphs, RELSYN-OPT runs in O(M·(L+N)), and is asymptotically optimum. RELSYN-OPT, because of its speed, applies to static & dynamic task allocation for an ultra-reliable distributed processing environment for which, until now, research has produced only suboptimal heuristic solutions.

Original languageEnglish (US)
Pages (from-to)603-613
Number of pages11
JournalIEEE Transactions on Reliability
Volume44
Issue number4
DOIs
StatePublished - 1995

Fingerprint

Military applications
Space applications
Computational complexity
Processing

Keywords

  • Heterogeneous multiprocessor system
  • task allocation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Safety, Risk, Reliability and Quality
  • Computer Graphics and Computer-Aided Design
  • Hardware and Architecture
  • Software

Cite this

Optimal Algorithms for Synthesis of Reliable Application-Specific Heterogeneous Multiprocessors. / Dasgupta, Aurobindo; Karri, Ramesh.

In: IEEE Transactions on Reliability, Vol. 44, No. 4, 1995, p. 603-613.

Research output: Contribution to journalArticle

@article{49362dd9c744490d9e556a12401a669e,
title = "Optimal Algorithms for Synthesis of Reliable Application-Specific Heterogeneous Multiprocessors",
abstract = "Fast and optimally-reliable application-specific multiprocessor-synthesis is critical in system-level design, especially in medical, automotive, space, and military applications. Previous work in multiprocessor-synthesis & task-allocation for performance & reliability requires exponential time, and therefore, is useful only for small examples. We present the first deterministic and provably-optimal algorithm (RELSYN-OPT) to synthesize real-time, reliable multiprocessors using a heterogeneous library of N processors and L link types. We prove that for a series-parallel graph with M subtasks and nested-depth d, the worst-case computational complexity of RELSYN-OPT is O(M·(L+N)·Nd). For tree-structured task graphs, RELSYN-OPT runs in O(M·(L+N)), and is asymptotically optimum. RELSYN-OPT, because of its speed, applies to static & dynamic task allocation for an ultra-reliable distributed processing environment for which, until now, research has produced only suboptimal heuristic solutions.",
keywords = "Heterogeneous multiprocessor system, task allocation",
author = "Aurobindo Dasgupta and Ramesh Karri",
year = "1995",
doi = "10.1109/24.475979",
language = "English (US)",
volume = "44",
pages = "603--613",
journal = "IEEE Transactions on Reliability",
issn = "0018-9529",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Optimal Algorithms for Synthesis of Reliable Application-Specific Heterogeneous Multiprocessors

AU - Dasgupta, Aurobindo

AU - Karri, Ramesh

PY - 1995

Y1 - 1995

N2 - Fast and optimally-reliable application-specific multiprocessor-synthesis is critical in system-level design, especially in medical, automotive, space, and military applications. Previous work in multiprocessor-synthesis & task-allocation for performance & reliability requires exponential time, and therefore, is useful only for small examples. We present the first deterministic and provably-optimal algorithm (RELSYN-OPT) to synthesize real-time, reliable multiprocessors using a heterogeneous library of N processors and L link types. We prove that for a series-parallel graph with M subtasks and nested-depth d, the worst-case computational complexity of RELSYN-OPT is O(M·(L+N)·Nd). For tree-structured task graphs, RELSYN-OPT runs in O(M·(L+N)), and is asymptotically optimum. RELSYN-OPT, because of its speed, applies to static & dynamic task allocation for an ultra-reliable distributed processing environment for which, until now, research has produced only suboptimal heuristic solutions.

AB - Fast and optimally-reliable application-specific multiprocessor-synthesis is critical in system-level design, especially in medical, automotive, space, and military applications. Previous work in multiprocessor-synthesis & task-allocation for performance & reliability requires exponential time, and therefore, is useful only for small examples. We present the first deterministic and provably-optimal algorithm (RELSYN-OPT) to synthesize real-time, reliable multiprocessors using a heterogeneous library of N processors and L link types. We prove that for a series-parallel graph with M subtasks and nested-depth d, the worst-case computational complexity of RELSYN-OPT is O(M·(L+N)·Nd). For tree-structured task graphs, RELSYN-OPT runs in O(M·(L+N)), and is asymptotically optimum. RELSYN-OPT, because of its speed, applies to static & dynamic task allocation for an ultra-reliable distributed processing environment for which, until now, research has produced only suboptimal heuristic solutions.

KW - Heterogeneous multiprocessor system

KW - task allocation

UR - http://www.scopus.com/inward/record.url?scp=0029534167&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029534167&partnerID=8YFLogxK

U2 - 10.1109/24.475979

DO - 10.1109/24.475979

M3 - Article

AN - SCOPUS:0029534167

VL - 44

SP - 603

EP - 613

JO - IEEE Transactions on Reliability

JF - IEEE Transactions on Reliability

SN - 0018-9529

IS - 4

ER -