Cedar Fortran and other vector and parallel Fortran dialects

Mark D. Guzzi, David A. Padua, Jay Hoeflinger, Duncan H. Lawrie

Research output: Contribution to journalArticle

Abstract

The introduction of vector processors and multiprocessors has caused the most dramatic changes in Fortran and its dialects. The emerging generation of supercomputers utilizes both vector processing and multiprocessing. The challenge is to provide language constructs and software tools that will allow the programmer to easily exploit the capabilities of the machine. This paper outlines the development of vector and multiprocessor language constructs in Fortran. Significant architectures, their languages, and optimizers are described. The paper concludes with a description of Cedar Fortran, the language for the Cedar Multiprocessor under development at the University of Illinois, Urbana-Champaign. Cedar is a hierarchical, shared-memory, vector multiprocessor. As such, its language, Cedar Fortran, contains many of the language features that are described for vector processors and multiprocessors.

Original languageEnglish (US)
Pages (from-to)37-62
Number of pages26
JournalThe Journal of Supercomputing
Volume4
Issue number1
DOIs
StatePublished - Mar 1 1990

Keywords

  • Cedar Fortran
  • Doacross
  • Doall
  • Fortran
  • block-parallel
  • language constructs
  • macrotasking
  • microtasking
  • multiprocessing
  • synchronization
  • unstructured tasking
  • vector processing

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Information Systems
  • Hardware and Architecture

Fingerprint Dive into the research topics of 'Cedar Fortran and other vector and parallel Fortran dialects'. Together they form a unique fingerprint.

Cite this