Synchronizing Large Systolic Arrays.

Abstract

Parallel computing structures consist of many processors operating simultaneously. If a concurrent structure is regular, as in the case of systolic array, it may be convenient to think of all processors as operating in lock step. Totally synchronized systems controlled by central clocks are difficult to implement because of the inevitable problem of clock skews and delays. An alternate means of enforcing necessary synchronization is the use of self-timed, asynchronous schemes, at the cost of increased design complexity and hardware cost. Realizing that different circumstances call for different synchronization methods, this paper provides a spectrum of synchronization models; based on the assumptions made for each model, theoretical lower bounds on clock skew are derived, and appropriate or best-possible synchronization schemes for systolic arrays are proposed. This paper represents a first step towards a systematic study of synchronization problems for large systolic arrays.

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1982
Accession Number
ADA123374

Entities

People

  • Allan L. Fisher
  • H. T. Kung

Organizations

  • Carnegie Mellon University

Tags

DTIC Thesaurus Topics

  • Computational Processes
  • Computing-Related Activities
  • Contract Administration
  • Contracts
  • Diffraction
  • Parallel Computing
  • Spectra

Fields of Study

  • Engineering

Readers

  • Parallel and Distributed Computing.
  • Positioning, Navigation, and Timing (PNT) Technology.
  • Theoretical Analysis.