Computing time is not a problem

PCI-II started in April 1996 and will finish in March 1998. The project is organized as a cluster of ten subprojects where each one develops an application for a parallel platform and deals with a different subject. The PCI-II project is managed by the Centre Europeu de Parallelisme de Barcelona (CEPBA) from Universitat Politècnica de Catalunya. PCI-II continues the work of PCI-PACOS (1994-1996), a preceding program from the European Commission aimed to improve the benefits and the performance of Spanish industries by using parallel technology.

PCI-II, instead, aspires to enhance the results of all kind of enterprises from all Europe. The achievements obtained by these projects will be continued by CEPBA TTN, a cluster of nine subprojects that started this year.

PCI-II has a cost of 6 millions of ECUS, 3.3 of them being financed by the European Comission. The only thing that the companies involved in PCI-II have in common is that all of them are trying to improve their competitiveness by using parallel technology.

The areas covered by the projects, as you can see below, are quite divers:

• AMODES. This project is based on the experimentation of a combination of advanced information technologies in the context of a DataWarehouse process of several Italian financial institutions.

• GLASSPAR aims to improve glass quality and, with this objective, the project has developed a parallel modelling code to aid the design process of the tin vat where glass sheets are produced.

• HIPEROAD develops a software system to improve the streamlining of the vehicles by choosing the most suitable external shape.

• MMIPPS means Multispectral and Multitemporal Image Process on Parallel Systems and it is an image-processing package that aims to boost computational intensive image-processing tasks.

• PARSAI focuses on the parallel simulation in the automotive industry and uses VECTIS, a commercial computational fluid dynamics (CFD) software package. CFD is used in simulations of flows in engine components.

• PARSAR. Synthetic Aperture Radar (SAR) obtains high resolution images of the Earth surface, and the PARSAR project is the porting of a sequential SAR code to a parallel architecture.

• PCECOWATER aims to develop a parallel software for studying the circulation of shallow waters in rivers, lakes,... The project also models other different kinds of environmental problems.

• RAPT-2 is a medical application of parallel HPC and allows an accurate calculation of dosage administered to cancer diseased.

• RETACO. The project aims to develop EXCHANGE, a parallel code that simulates the ground water flow and the behaviour of chemicals in water.

• STAMPAR aims to apply parallel computation techniques to a serial simulation code (named Stampack) for sheet stamping processes.

All these projects have been put together in order to maximize their impact in society. A unique subproject would not have obtained the same effect among enterprises and industry. The message that PCI-II wants to transmit is clear: a great part of the problems that an enterprise may think about can be solved by parallel technology, so the industrial sector only has to care about the subject of study but not about the computing time the work is going to spend. This time can be reduced substantialy by using parallel tecnology.

CEPBA is satisfied with the results obtained in PCI-II. The most important achievement is that enterprises have pushed improvements in their ways of working and in facing up to their bussines. Some of the subprojects developed in PCI-II are close to commercial exploitation, and this is a second achievement CEPBA is proud of. Exploitation is one of the key points where we have put emphasis, said Jesús Labarta, director of CEPBA.