An integrated framework for rapid system prototyping and automatic code distribution

Abstract

Rapid prototyping of parallel systems is of interest to quickly produce a parallel prototype. The emergence of distributed systems technology has enabled one to develop software systems distributed over large networks. Rapid prototyping must deal with real parallelism over a set of processors, either closely or loosely coupled. We describe an extension of the CPN/TAGADA project to manage distributed code generation over a set of CPU. To achieve a mapping of components over the target architecture, both hardware and software have to be described. We expose our technique and apply it to a multi producer and consumer example that is studied for several communication strategies.<>