Towards a Bulk-Synchronous Distributed Shared Memory Programming Environment for Grids

* Final gross prices may vary according to local VAT.

Get Access

Abstract

The current practice in grid programming uses message passing, which unfortunately leads to code that is difficult to understand, debug and optimize. Hence, for grids to become commonly accepted, also as general-purpose parallel computation platforms, suitable parallel programming environments need to be developed.

In this paper we propose an approach to realize a distributed shared memory programming environment for computational grids called GridNestStep, by adopting NestStep, a structured parallel programming language based on the Bulk Synchronous Parallel model of parallel computation.