Solving irregular problems with High Performance Fortran

Abstract

High Performance Fortran (HPF) provides a high-level interface for data-parallel scientific applications, while delegating to the compiler the task of generating an explicitly parallel message-passing program. The main objective of this paper is to study the applicability of the language to a class of irregular scientific applications. After giving an outline of developments that led to HPF and shortly explaining its major features, we discuss a variety of approaches for solving multiblock problems and applications dealing with unstructured meshes. We show that the efficient solution of these problems does not only need the full range of the HPF approved extensions, but also requires additional features such as the explicit control of communication schedules and support for value-based alignment.