Fujitsu Will Supply New Supercomputer System to University of Tokyo
OREANDA-NEWS. August 24, 2012. Fujitsu announced that it has received an order for a new supercomputer from the Institute for Solid State Physics (ISSP) at The University of Tokyo. The new system is scheduled to begin operations next spring.
The new system, a Fujitsu PRIMEHPC FX10 supercomputer, will be comprised of four racks, with a total of 384 nodes for a theoretical computational performance of 90.8 teraflops. The system will primarily be used to support nationwide research conducted collaboratively by solid state science researchers, in addition to being employed in research and applications development for the 2nd field of the K computer's strategic program, namely, "New Materials and Energy Creation." Researchers in this field aim to further the development of new materials and to produce key technologies that lead to the creation of new energy, while conducting R&D that will help bring about a sustainable society. The new system is expected to spur dramatic advances in the field of solid state sciences research and contribute to the development of high-performance devices, high-efficiency fuel cells, and new drugs.
Background to the Deployment of the New System
Since its establishment in 1957, ISSP has served as
To facilitate the development of applications for use in the K computer's strategic program and other areas, ISSP selected the Fujitsu PRIMEHPC FX10 supercomputer, which is highly compatible with the K computer, having the latest ultra-parallel computing technologies. As a result, researchers will be able to develop and fine-tune programs in an even more practical manner, thereby helping to facilitate breakthrough achievements in solid state science research while training the next generation of researchers who will lead the field of computational materials science.
Overview of the New System
Combining high performance, scalability, and reliability with superior energy efficiency, PRIMEHPC FX10 further enhances the Fujitsu technology used in the K computer.
The new system will use a PRIMEHPC FX10 supercomputer comprised of 4 racks and 384 nodes, thereby achieving a total theoretical computational performance of 90.8 teraflops. For its HPC middleware, the system will employ Technical Computing Suite(4) for peta-scale systems, together with 12 PRIMERGY series servers as login nodes and management servers. ETERNUS storage systems, combined for a total capacity of 126.9 terabytes, will be deployed for storage. The file system will be constructed using the high-capacity, high-performance and highly reliable FEFS(5) distributed file system.
PRIMEHPC FX10 received high acclaim from ISSP for its high performance and reliability, as well as its superior energy efficiency that can keep up with today's energy requirements, and the ultra-high performance and usability of FEFS.
Future Prospects
Using the new system, ISSP will strive to train researchers who can develop applications for the K computer, together with accelerating research activities conducted on the K computer. This, in turn, is expected to lead to dramatic advances in solid state science and contribute to the development of high-performance devices, high-efficiency fuel cells, and new drugs.
Комментарии