Notification of ACI-GRID 2002



Speech by Mr Roger-Gérard Schwartzenberg, Minister of Research Paris, February 28, 2001 Launching of the ACI "Globalization of computer resources and data" (GRID)At the IST (Information Society Technologies) 2000 symposium in Nice,

I announced a series of measures aimed at developing state-funded research in information and communications science and technology. One of the measures put forward was the setting up of an "Incentive Concerted Action" (ACI) devoted to megadatabases and distributed computing. I am pleased to see that, three months later, we are ready to launch this "ACI - Globalization of computer resources and data" (GRID). Thank you for agreeing to steer it under the guidance of Michel Cosnard. The GRID ACI will receive the backing of 15 million French francs from the Ministry of Research, provided by the "Fonds National de la Science" (National Science Fund). It was important for this initiative to be taken rapidly. Firstly, because the notion of data storage and of computing resources is changing. Secondly, because the French system in this field has always been at the forefront of research and constitutes a springboard for new breakthroughs. Finally, because the issues raised by "distributed computing" require the coordination of researchers from various sectors - which is precisely what is facilitated by the "ACI" scheme. *** Needs for computation and storage resources are growing fast, while networking such resources will soon create new opportunities. As you know, we are currently experiencing very high growth in needs in terms both of capacity to store data and also of computation power to process it. For example, this applies to nuclear physics, for simulating trials, to biology, for sequencing the genome, to meteorology, with forecasting models, and to particle physics, for searching for the minutest components of matter - I would remind you that the Large Hadron Collider (LHC) is to be brought into service at CERN in 2005. But beyond the world of research, industry and business are experiencing an increase in needs, and small businesses must now have access to increasingly high performance tools. One of the first paths that we are following is naturally the continuous reinforcement of the capacities of the main computing centers available to our research teams. The CINES ("Centre Informatique National de l'Enseignement Supérieur" or "National Computer Center for Higher Education"), the CEA (Atomic Energy Commission) and, at the CNRS, the IDRIS ("Institut du Développement et des Ressources en Informatique" or "Institute for Development and Resources in Intensive Scientific Computing") thus enjoy facilities that are among the world's best. A few weeks back, we announced that IDRIS is commissioning a new supercomputer that will be installed by the end of year, and that will offer power of 1.3 teraflops (1,300 billion floating-point operations per second), and storage capacity of 900 gigabytes (nine billion bytes), for a cost of 54 million francs. These centers are supplemented by smaller-scale local facilities, that are often located in universities. At the same time, the rise of the interconnection networks, with the increase Renater's capacities more than sixteen-fold in 2001, and the setting up of the GEANT project at the European level, offers researchers access to these capacities from their laboratories. More generally, the networking of existing and future resources offers new prospects that we must explore, through the computation and storage grid. The project of a Worldwide Grid, which is to computation what the Internet is to communications, promises almost unlimited capacity: all the resources would be accessible as and when required. Above all, it offers the possibility of optimizing the available resources, which are usually financed from public funds. *** Although the network nodes (computers and servers) already exist, and although the networks that will convey the computation are being built, many problems still remain unsolved. How will this project be implemented? Several scientific consortiums in Europe and in the United States have been set up to design the missing software layers, and to improve their performance. By launching this ACI, we wish to boost the involvement of French teams in meeting this challenge which requires considerable coordination efforts. That is why the scientific council of this ACI is made up of computer science specialists and of specialists from all the communities of users of these new techniques. The questions raised today essentially concern the following points: The distribution of tasks between various computers, which should be clearly understood by the user. The basic computing code, which should be as simple and user-friendly as possible. The necessity for absolute security: the interchange of data over the networks and the sharing of resources must under no circumstances lead to the loss or interception of a computation result. The integration of various types of computer, from the supercomputer to the personal computer. The choice of the communications protocol. Your mission will thus be to make the contribution from French research teams "take off" at this critical time by supporting the teams that are already active, by attracting new players, and by fostering meetings between designers and users of new solutions. It will also be necessary to ensure the French effort fits into an international, and in particular European, context. *** To achieve the objectives of this ACI, several forms of action should be considered: call for project proposals, support for young researchers, organizing symposia and summer schools, sending French researchers abroad and hosting foreign specialists in France, etc. I would like these actions to be launched in the coming weeks, so that the research teams may receive the chose form of support by this summer. In this manner, I hope, we will have made sufficient headway to enable our research teams to take up a leading position in this new deal. Indeed, these developments concern the entire scientific community and their repercussions will be felt both in the research world and in industry.