December 09, 2005
To aid global environment research, Silicon Graphics and SGI Japan Limited recently installed a new supercomputing system at the Institute of Fluid Science, Tohoku University. The new system was built using SGI server, visualization and storage technology to enable researchers to achieve more accurate results and shorter time to solution.
The Institute of Fluid Science, Tohoku University is a research facility that specializes in fluid science - the flow of material substance as well as heat, energy and information. As a pioneer, the institute has achieved a number of findings by conducting large-scale numerical simulations essential to fluid science research. The institute selected SGI Japan's proposal for its system based on the requirements for the new supercomputer and on SGI's experience in the supercomputing field in Japan.
The new supercomputer, shipped this summer, is comprised of an SGI Altix 3000 scalar parallel server based on the 64-bit Linux OS, a vector parallel NEC computer, a scalable Silicon Graphics Prism visualization system, external secondary storage systems and data archive systems, all of which are interconnected via high-speed network, enabling sharing of large files with the SGI InfiniteStorage CXFS shared filesystem and the NEC GFS global file system attached to the Storage Area Network.
The core of the new system is one of the fastest computers to have been installed in Japan for fluid science researches. Based on the SGI Altix system, the supercomputer will contain 1,024 Intel Itanium 2 processors and 12 TB of main memory in a large-scale shared-memory configuration, or single system image (SSI). The Institute of Fluid Science has installed a Silicon Graphics Prism visualization system, containing 256 processors, eight ATI graphics processors and two TB of main memory, which will be used to accelerate the analysis of results generated on the Altix system. The new system will also provide one petabyte of storage, one of the largest storage systems in Japan.
In April, 2003, the institute expanded by establishing a next-generation Fluid Integration Research Center where scientists promote fluid science researches in various advanced science and engineering fields by means of a new research methodology integrating its experimental and computational methods. To analyze and control increasingly complex and diversified fluid problems, the institute decided to replace its existing supercomputer with a new system specifically designed for-next generation fluid integration research for even more accurate results and shorter time to solution.
The institute also needs to enhance its interface server and visualization server for real-time integration of experimental and simulation results and 3D visualization to solve complex flow problems seen in actual phenomena through its new methodology. The Silicon Graphics Prism system will be used to understand the results generated by the next generation of fluid simulations. The institute will also use SGI's Visual Area Networking technology to deliver interactive, visual results directly to researchers' desktops independent of their location.
The disk and tape storage systems will provide one petabyte of capacity, sufficient to process and store the research data. By connecting both the supercomputers and visualization system to the SAN based on CXFS, researchers will be able to share one set of storage for their entire research process. This accelerates fluid research by eliminating time-consuming data copies and simplifying storage management.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?