PAMPAProgramming environment of massively parallel architecturesContext and objectivesPampa is a common project with CNRS and the University of Rennes 1. The development of computer networks allowing the interconnection of machines is increasing. Therefore the questions raised by the software construction for these systems are of a great topicality. Even if spectacular progress were accomplished in the methods of software engineering, the parallel and distributed character of the considered software continues to pose difficult problems of programming. The most sensitive question in our opinion is the control of the quality of the software, i.e. the control of the conditions of sound correct operation. The control of the development passes by the reinforcement of the activities of design, validation and test. From the design point of view, priority is given to the object-oriented design environments and the construction of customized frameworks specialized for communicating systems and integrating validation tools. From the point of view of the validation, the idea is to reinforce the impact of formal methods and tools to allow the debugging of specifications and the generation of tests for the distributed codes. The approach of the Pampa project consists in contributing to the development of new software technologies by the study of formal models for protocols and the invention of associated software tools. We privilege the design of automatic tools allowing to contribute to the tasks of design, verification, code generation and test of real programs. These tools have vocation to be distributed in the academic and/or industrial medium. The development of the models and tools is carried out within the framework of distributed applications located mainly in the field of the software for telecommunications.
Research axes
International and industrial partnerships
Last modified
2008/05/19 10:06
|
Scientific leaderAddress
IRISA - Campus universitaire de Beaulieu - 35042 Rennes Cedex |