A Parallel Object-Oriented Framework for Stencil Algorithms

Report
Authors:Karpovich, John, Department of Computer ScienceUniversity of Virginia Judd, Matthew, Department of Computer ScienceUniversity of Virginia Strayer, W, Department of Computer ScienceUniversity of Virginia Grimshaw, Andrew, Department of Computer ScienceUniversity of Virginia
Abstract:

We present an object-oriented framework for constructing parallel implementations of stencil algorithms. This framework simplifies the development process by encapsulating the common aspects of stencil algorithms in a base stencil class so that application-specific derived classes can be easily defined via inheritance and overloading. In addition, the stencil base class contains mechanisms for parallel execution. The result is a high-performance, parallel, application-specific stencil class. We present the design rationale for the base class and illustrate the derivation process by defining two sub-classes, an image convolution class and a PDE solver. The classes have been implemented in Mentat, an object-oriented parallel programming system that is available on a variety of platforms. Performance results are given for a network of Sun SPARCstation IPCs.

Rights:
All rights reserved (no additional license for public reuse)
Language:
English
Source Citation:

Karpovich, John, Matthew Judd, W Strayer, and Andrew Grimshaw. "A Parallel Object-Oriented Framework for Stencil Algorithms." University of Virginia Dept. of Computer Science Tech Report (1993).

Publisher:
University of Virginia, Department of Computer Science
Published Date:
1993