Automatic vectorisation for high level languages based on an expert system

  • Thomas Brandes
System Software & Programming Language Aspects (Session 4.1)
Part of the Lecture Notes in Computer Science book series (LNCS, volume 237)


New hardware architecture use vector operations and parallelism for high speed computation; existing software has mostly to be rewritten. The automation of the adaptation of existing software to the new hardware is unsatisfactory, especially when using high level languages, as the vectorisation and parallelisation process needs a knowledge of the hardware, of convenient transformations and of semantic properties of the program to take advantage of the hardware possibilities. The expert system SAVER will be the basis to accelerate and automate thoroughly the vectorisation process. The most interesting features of the Saver system will be presented in this paper.


Expert System Semantic Property Hardware Architecture High Level Language Program Transformation 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

8. Bibliography

  1. [1]
    Hwang,K.; Shun-Piao,S.; Noi,L.M:Vector Computer Architecture and Processing Techniques; Advances in Computers, Vol. 20Google Scholar
  2. [2]
    Kowalik,S.J.: High Speed Computation; Springer Verlag 1984; Series F: Computer and Systems Sciences, Vol. 7, Part 1Google Scholar
  3. [3]
    Perrott,R.H.; Crookes,D.; Milligan,P.; Purdy,M.: A Compiler for an Array and Vector Processing Language; IEEE — Transactions on SE, Vol. 11, No.5, May 1985Google Scholar
  4. [4]
    Li,K.C.; Schwetmann,H.: VectorC: A Vector Processing Language; Journal of Parallel and Distributed Computing 2Google Scholar
  5. [5]
    Ehlich,H.: PASCALV: Der Pascal-Compiler für den Vektorrechner CYBER 205; Bochumer Schriften zur Parallelen DatenverarbeitungGoogle Scholar
  6. [6]
    Völksen,G.; Wehrum, P.: Transition to ADA for Super Computers; ADA-Europe Conference Edinburgh, May 1986; Cambridge University PressGoogle Scholar
  7. [7]
    Peterson,W.P.: Vector Fortran for Numerical Problems on CRAY-1; Comm. of the ACM: November 1983, Vol. 26, No. 11Google Scholar
  8. [8]
    Siemens: Vector Processor System VP100,VP200; Siemens 7-800 System DescriptionGoogle Scholar
  9. [9]
    Nagel,W.: Ein Preprocessor zur Unterstützung vektorisierender Compiler; Diplomarbeit am Institut für Allgemeine Elektrotechnik und DV-Systeme der Rheinisch-Westfälischen Technischen Hochschule AachenGoogle Scholar
  10. [10]
    Lamport,L.: The Parallel Execution of DO Loops; Comm. of the ACM: Feb. 1974, Vol. 17, No.2Google Scholar
  11. [11]
    Kogge, P.M.: Parallel Solution of Recurrence Problems; IBM J.Res. Development, March 1974Google Scholar
  12. [12]
    Lee,G.; Kruskal,C.; Kuck,D.: An Empirical Study of Automatic Restructuring of Nonnumerical Programs for Parallel Processors; IEEE Transactions on Computers Vol C-34, No 10, Oct. 85Google Scholar
  13. [13]
    Allen,J.R.: Kennedy,K.: A Parallel Programming Environment; IEEE Software, July 1985Google Scholar
  14. [14]
    Teitelman, W.: Interlisp Reference Manual; Technical Report Xerox PARCGoogle Scholar
  15. [15]
    Bobrow,D.; Stefik,M.: The LOOPS Manual; Technical Report Xerox PARCGoogle Scholar
  16. [16]
    Steels, L.: Design Requirements for Knowledge Base Representation Systems; 8th German Workshop on Artifical Intelligence Wingst/Stade 1984; Springer VerlagGoogle Scholar
  17. [17]
    E.Rich: Artifical Intelligence; Mc Graw-Hill Book CompanyGoogle Scholar
  18. [18]
    Stefik, M.; Bobrow, D.G.; Mittal, S.; Conway, L.: Knowledge Programming in LOOPS: Report on an experimental course; Artifical Intelligence 4,3 (Fall 1983), pp. 3–14Google Scholar
  19. [19]
    Fikes,R.; Kehler,T.: The Role of Frame-Based Representation in Reasoning; Comm. of the ACM: September 1985, Vol. 28, No. 9Google Scholar
  20. [20]
    Sommer,M.: Vektorisierung von Algorithmen; Siemens Forschungs-und Entwicklungs-berichte, Heft 5/86Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1986

Authors and Affiliations

  • Thomas Brandes
    • 1
  1. 1.FB Mathematik/Informatik, Philipps-UniversitätMarburg, Lahnberge

Personalised recommendations