Design of Optimal Systolic Algorithms for the Transitive Closure Problem

Dilip Sarkar, Amar Mukherjee

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We present new optimal systolic algorithms for the transitive closure problem on ring and linear array of processors. The data dependency of the Warshal-Floyd algorithm is exploited to obtain highly pipelined parallel algorithms. One of the algorithms is asymptotically seven times cost-effective than the existing algorithms for computing transitive closure problems. We introduce a new expository device, called the RCT diagram, that depicts simultaneous flow of data and computation for parallel algorithms.

Original languageEnglish (US)
Pages (from-to)508-512
Number of pages5
JournalIEEE Transactions on Computers
Volume41
Issue number4
DOIs
StatePublished - Jan 1 1992

Fingerprint

Systolic arrays
Transitive Closure
Parallel algorithms
Parallel Algorithms
Randomized Controlled Trial
Data Dependency
Linear Array
Diagram
Ring
Design
Costs
Computing

Keywords

  • Data dependency
  • linear arrays
  • parallel algorithms
  • parallel machines
  • RCT diagram
  • systolic arrays
  • transitive closure problem
  • VLSI
  • WSI

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Software
  • Hardware and Architecture
  • Computational Theory and Mathematics

Cite this

Design of Optimal Systolic Algorithms for the Transitive Closure Problem. / Sarkar, Dilip; Mukherjee, Amar.

In: IEEE Transactions on Computers, Vol. 41, No. 4, 01.01.1992, p. 508-512.

Research output: Contribution to journalArticle

@article{a2a77b3735434e938af463bc9383f300,
title = "Design of Optimal Systolic Algorithms for the Transitive Closure Problem",
abstract = "We present new optimal systolic algorithms for the transitive closure problem on ring and linear array of processors. The data dependency of the Warshal-Floyd algorithm is exploited to obtain highly pipelined parallel algorithms. One of the algorithms is asymptotically seven times cost-effective than the existing algorithms for computing transitive closure problems. We introduce a new expository device, called the RCT diagram, that depicts simultaneous flow of data and computation for parallel algorithms.",
keywords = "Data dependency, linear arrays, parallel algorithms, parallel machines, RCT diagram, systolic arrays, transitive closure problem, VLSI, WSI",
author = "Dilip Sarkar and Amar Mukherjee",
year = "1992",
month = "1",
day = "1",
doi = "10.1109/12.135564",
language = "English (US)",
volume = "41",
pages = "508--512",
journal = "IEEE Transactions on Computers",
issn = "0018-9340",
publisher = "IEEE Computer Society",
number = "4",

}

TY - JOUR

T1 - Design of Optimal Systolic Algorithms for the Transitive Closure Problem

AU - Sarkar, Dilip

AU - Mukherjee, Amar

PY - 1992/1/1

Y1 - 1992/1/1

N2 - We present new optimal systolic algorithms for the transitive closure problem on ring and linear array of processors. The data dependency of the Warshal-Floyd algorithm is exploited to obtain highly pipelined parallel algorithms. One of the algorithms is asymptotically seven times cost-effective than the existing algorithms for computing transitive closure problems. We introduce a new expository device, called the RCT diagram, that depicts simultaneous flow of data and computation for parallel algorithms.

AB - We present new optimal systolic algorithms for the transitive closure problem on ring and linear array of processors. The data dependency of the Warshal-Floyd algorithm is exploited to obtain highly pipelined parallel algorithms. One of the algorithms is asymptotically seven times cost-effective than the existing algorithms for computing transitive closure problems. We introduce a new expository device, called the RCT diagram, that depicts simultaneous flow of data and computation for parallel algorithms.

KW - Data dependency

KW - linear arrays

KW - parallel algorithms

KW - parallel machines

KW - RCT diagram

KW - systolic arrays

KW - transitive closure problem

KW - VLSI

KW - WSI

UR - http://www.scopus.com/inward/record.url?scp=0026853564&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026853564&partnerID=8YFLogxK

U2 - 10.1109/12.135564

DO - 10.1109/12.135564

M3 - Article

AN - SCOPUS:0026853564

VL - 41

SP - 508

EP - 512

JO - IEEE Transactions on Computers

JF - IEEE Transactions on Computers

SN - 0018-9340

IS - 4

ER -