|
Press Release
Armed With 13-Fold Increase In Compute Power, HLRN Inaugurates New SGI Supercomputer First Phase of Multi-Year Deployment to Help Scientific Community Solve 'The Most Challenging Problems of Our Time' SUNNYVALE, Calif. and MUNICH (July 3, 2008)— Ushering in a new future for Germany's scientific community, German officials today inaugurated a powerful new supercomputer equipped with systems from SGI (NASDAQ: SGIC).The North German Supercomputing Alliance (HLRN) maintains the new system in two separate cities, Berlin and Hanover. Both sites have installed identical SGI® supercomputers, and although they are located 155 miles apart, they can be combined to attack large-scale problems via a high-speed network. Those problems include complex simulations and data-intensive processes in physics, materials research, computational fluid mechanics, astrophysics, chemistry, and earth sciences. Known as HLRN-II, the combined SGI installation:
When the second phase of the project is completed at the end of 2009, HLRN-II will be more than four times as powerful as it is today, generating an anticipated peak performance of 312 trillion operations per second (Teraflops). Meanwhile, HLRN's choice of energy-efficient SGI solutions will minimize the system's environmental impact on operation centers in Berlin and Hanover. "Having started the conceptual design process only four months ago, the first phase of HLRN-II will soon be available to scientists," said Prof. Gabriele von Voigt, managing director of the Regional Computer Centre of Lower Saxony, where the Hanover half of HLRN-II is located at Leibniz University's Regionales Rechenzentrum für Niedersachsen (RRZN). "Researchers working at northern Germany's state-run facilities will use this resource to tackle some of the most challenging problems of our time." The Berlin half of HLRN-II is located at Konrad-Zuse-Zentrum für Informationstechnik (ZIB). "We are eager to see how our scientists will turn the increased computing power of HLRN-II and its ability to manage vast amounts of data efficiently into groundbreaking new insights," said Prof. Alexander Reinefeld, director of Computer Science at ZIB. "A world-class high-performance system requires optimized applications to achieve extraordinary results, so the first step is for our consultants to help the users in adjusting their codes in engineering, earth sciences, physics and computational chemistry." Both systems are connected via "HLRN Link," a network-based environment that provides researchers with a shared, massively parallel computing system with common access to enormous amounts of data. Using HLRN Link, researchers can also store and manage the immense volumes of data needed to generate ever-larger scientific models and simulations. To achieve this HRLN-II has increased its total SGI InfiniteStorage infrastructure by another 1.15 Petabytes, and will double that capacity in 2009. "Around the world, society has tasked scientists with improving quality of life by rapidly finding new insights and discoveries to provide reliable solutions to some of our most daunting problems. HLRN provides the scientist with a world-class high-performance computing environment in support of their research," said Robert Übelmesser, Managing Director of Silicon Graphics GmbH in Germany. "Not only are SGI systems designed to help researchers achieve new breakthroughs, but they also enable data centers to lower their energy consumption and cooling costs. With SGI, customers like HLRN achieve the highest performance while minimizing their environmental impact." Multimedia Resources:
Additional Product Information:
RSS Feeds: About HLRN SGI - Innovation for Results™ © 2008 SGI. All rights reserved. SGI, Altix, the SGI cube and the SGI logo are registered trademarks of SGI in the United States and/or other countries worldwide. All other trademarks mentioned herein are the property of their respective owners * Research related to this work has led to publications in Physical Review Letters [98, 168301 (2007)], Physical Review E [72, 031808 (2005)] and Macromolecules [39, 3007-3018 (2006) and 40, 4568 (2007)]. This work has been carried out by the group of "Polymers and non-Crystalline Materials" (DIPC, CSIC-UPV, San Sebastian, Spain) led by Professor Colmenero. | |