A parallel structure for static iterative transformation algorithms

Abstract

Parallel functional programming could be made more efficient if the paradigm used was clearly identified. A paradigm can be expressed by defining a template called a parallel algorithmic structure. As a case study, this paper addresses the issue of defining a parallel structure for the static iterative transformation paradigm. It shows how a specification can be successively turned into a sequential functional program, then into a parallel program, and finally into a program that maps on a specific parallel architecture.