Publication View

Generalized Unimodular Loop Transformations for Distributed Memory Multiprocessors (1995)

Abstract
In this paper, we present a generalized unimodular loop transformation as a simple, systematic and elegant method for partitioning the iteration spaces of nested loops for execution on distributed memory multiprocessors. We present a methodology for deriving the transformations that internalize multiple dependences in a multidimensional iteration space without resulting in a deadlocking situation. We then derive the general expression for the bounds of the transformed loops in terms of the bounds of the original space and the transformation matrix elements. Keywords: Parallelizing Compilers, Restructuring Transformations, Loop Partitioning, Iteration Spaces, Dependence Vectors. 1 Introduction The great difficulty of programming distributed memory multiprocessors has motivated a spate of research activity in the automatic source level restructuring of sequential programs for execution on such machines. Many among these efforts have addressed the problem of partitioning loops since, ne...

Publication details
Download http://citeseer.ist.psu.edu/304941.html
Source ftp://ftp.cs.toronto.edu/pub/parallel/Kumar_Kulkarni_ICPP91.ps.Z
Publisher unknown
Contributors The Pennsylvania State University CiteSeer Archives
Repository CiteSeer (United States)
Keywords K G Kumar,D Kulkarni,A Basu Generalized Unimodular Loop Transformations for Distributed Memory Multiprocessors
Language Englisch
Relation oai:CiteSeerPSU:108359, oai:CiteSeerPSU:100681