The Forschungszentrum Jülich, Germany's leading HPC center and the instigator of the JuRoPA project, aimed at accelerating the development of high performance cluster computing in Europe, has ordered from Bull a supercomputer with a performance of 200 Teraflops, to be deployed at the beginning of 2009 and integrated within the Jülich's data center network.
"Science and industry increasingly rely and profit from simulations on computers of the highest performance class," explained Prof. Thomas Lippert, Director of the Jülich Supercomputing Centre. “To support these growing demands Jülich has chosen Bull as prime contractor for the JuRoPA project which also includes Sun, ParTec and Intel. We consider Bull to be best suited to bring together the leading technologies in a partnership to create a most energy and performance effective general-purpose supercomputer for the wide range of simulation problems tackled in Jülich.”
The JuRoPA project (which stands for "Jülich Research on Petaflops Architectures") was set up by the Forschungszentrum Jülich to investigate emerging cluster technologies and achieve a new class of cost-efficient supercomputers for peta-scale computing. Explored technologies include hardware based on Intel® Xeon® processors and innovative extreme bandwidth low-latency network technologies, as well as a new class of HPC middleware from ParTec. Intel, ParTec, and Sun are contributors to the project, with Bull taking on overall responsibility.
A Bull architecture integrating best-of-breed technologies in a production-level system
Jülich's new supercomputer is a cluster combining Bull NovaScale servers, Sun blade servers, all based on future Intel® Xeon® processors, and complete HPC cluster software provided by ParTec. This is combined with a high-performance input-output (I/O) system based on a Sun ZFS/Lustre file system, guaranteeing end-to-end data integrity.
Bull supplies the know-how to integrate these advanced technologies while ensuring that the new supercomputer is an extremely effective tool with the production-level standards of a world-class HPC center.
The new system increases the currently available computing power by a factor of 20. It opens the door to advanced research projects, offering researchers the possibility to work on today's major challenges in scientific research including, for example, energy management, novel materials and climatology.
"Bull is very proud to be selected by the Forschungszentrum Jülich, which is one of the world’s largest research centers dedicated to high performance computer simulation" said Philippe Miltin, Vice President Bull Products and Systems. "It is also a major step in the involvement of Bull in computer simulation, and an acknowledgement of Bull's high-level know-how both as a technology provider and as an architect of complex and flexible solutions, in one of the most dynamic markets today".
With this contract, Bull is signaling that it can meet the European challenge of high-performance computer simulation and act as a major industrial player in this area, particularly through partnerships with some of the leading organizations. Having anticipated several years ago the current evolution in computer simulation technologies, Bull is today at the forefront of this discipline, working alongside the best. This represents a considerable strength for the European industry, and will benefit all Bull customers.