On high-speed computing with a programmable linear array

Peizong Lee, Zvi M. Kedem

Research output: Contribution to journalArticle

Abstract

It has been observed by many researchers that systolic arrays are very suitable for certain high-speed computations. Using a formal methodology, we present a design for a single simple programmable linear systolic array capable of solving large numbers of problems drawn from a variety of applications. The methodology is applicable to problems solvable by sequential algorithms that can be specified as nested for-loops of arbitrary depth. The algorithms of this form that can be computed on the array presented in this paper include 25 algorithms dealing with signal and image processing, algebraic computations, matrix arithmetic, pattern matching, database operations, sorting, and transitive closure. Assuming bounded I/O, for 18 of those algorithms the time and storage complexities are optimal, and therefore no improvement can be expected by using dedicated special-purpose linear systolic arrays designed for individual algorithms. We also describe another design which, using a sufficient large local memory and allowing data to be preloaded and unloaded, has an optimal processor/time product.

Original languageEnglish (US)
Pages (from-to)223-249
Number of pages27
JournalJournal of Supercomputing
Volume4
Issue number3
DOIs
StatePublished - Sep 1990

Fingerprint

Linear Array
Systolic Array
High Speed
Systolic arrays
Computing
Matrix Computation
Transitive Closure
Methodology
Sequential Algorithm
Pattern Matching
Sorting
Pattern matching
Signal Processing
Image Processing
Sufficient
Signal processing
Image processing
Arbitrary
Data storage equipment
Design

Keywords

  • algebraic computations
  • algorithm transformations
  • database operations
  • matrix arithmetic
  • nested loop algorithms
  • Parallel processing
  • pattern matching
  • Programmable systolic array
  • signal and image processing
  • sorting
  • transitive closure
  • VLSI

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Electrical and Electronic Engineering
  • Hardware and Architecture
  • Computational Theory and Mathematics

Cite this

On high-speed computing with a programmable linear array. / Lee, Peizong; Kedem, Zvi M.

In: Journal of Supercomputing, Vol. 4, No. 3, 09.1990, p. 223-249.

Research output: Contribution to journalArticle

@article{61935e6e0b1d494ba9108858a52b41af,
title = "On high-speed computing with a programmable linear array",
abstract = "It has been observed by many researchers that systolic arrays are very suitable for certain high-speed computations. Using a formal methodology, we present a design for a single simple programmable linear systolic array capable of solving large numbers of problems drawn from a variety of applications. The methodology is applicable to problems solvable by sequential algorithms that can be specified as nested for-loops of arbitrary depth. The algorithms of this form that can be computed on the array presented in this paper include 25 algorithms dealing with signal and image processing, algebraic computations, matrix arithmetic, pattern matching, database operations, sorting, and transitive closure. Assuming bounded I/O, for 18 of those algorithms the time and storage complexities are optimal, and therefore no improvement can be expected by using dedicated special-purpose linear systolic arrays designed for individual algorithms. We also describe another design which, using a sufficient large local memory and allowing data to be preloaded and unloaded, has an optimal processor/time product.",
keywords = "algebraic computations, algorithm transformations, database operations, matrix arithmetic, nested loop algorithms, Parallel processing, pattern matching, Programmable systolic array, signal and image processing, sorting, transitive closure, VLSI",
author = "Peizong Lee and Kedem, {Zvi M.}",
year = "1990",
month = "9",
doi = "10.1007/BF00127833",
language = "English (US)",
volume = "4",
pages = "223--249",
journal = "Journal of Supercomputing",
issn = "0920-8542",
publisher = "Springer Netherlands",
number = "3",

}

TY - JOUR

T1 - On high-speed computing with a programmable linear array

AU - Lee, Peizong

AU - Kedem, Zvi M.

PY - 1990/9

Y1 - 1990/9

N2 - It has been observed by many researchers that systolic arrays are very suitable for certain high-speed computations. Using a formal methodology, we present a design for a single simple programmable linear systolic array capable of solving large numbers of problems drawn from a variety of applications. The methodology is applicable to problems solvable by sequential algorithms that can be specified as nested for-loops of arbitrary depth. The algorithms of this form that can be computed on the array presented in this paper include 25 algorithms dealing with signal and image processing, algebraic computations, matrix arithmetic, pattern matching, database operations, sorting, and transitive closure. Assuming bounded I/O, for 18 of those algorithms the time and storage complexities are optimal, and therefore no improvement can be expected by using dedicated special-purpose linear systolic arrays designed for individual algorithms. We also describe another design which, using a sufficient large local memory and allowing data to be preloaded and unloaded, has an optimal processor/time product.

AB - It has been observed by many researchers that systolic arrays are very suitable for certain high-speed computations. Using a formal methodology, we present a design for a single simple programmable linear systolic array capable of solving large numbers of problems drawn from a variety of applications. The methodology is applicable to problems solvable by sequential algorithms that can be specified as nested for-loops of arbitrary depth. The algorithms of this form that can be computed on the array presented in this paper include 25 algorithms dealing with signal and image processing, algebraic computations, matrix arithmetic, pattern matching, database operations, sorting, and transitive closure. Assuming bounded I/O, for 18 of those algorithms the time and storage complexities are optimal, and therefore no improvement can be expected by using dedicated special-purpose linear systolic arrays designed for individual algorithms. We also describe another design which, using a sufficient large local memory and allowing data to be preloaded and unloaded, has an optimal processor/time product.

KW - algebraic computations

KW - algorithm transformations

KW - database operations

KW - matrix arithmetic

KW - nested loop algorithms

KW - Parallel processing

KW - pattern matching

KW - Programmable systolic array

KW - signal and image processing

KW - sorting

KW - transitive closure

KW - VLSI

UR - http://www.scopus.com/inward/record.url?scp=0040822522&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0040822522&partnerID=8YFLogxK

U2 - 10.1007/BF00127833

DO - 10.1007/BF00127833

M3 - Article

AN - SCOPUS:0040822522

VL - 4

SP - 223

EP - 249

JO - Journal of Supercomputing

JF - Journal of Supercomputing

SN - 0920-8542

IS - 3

ER -