HPC NEWS
Contents HPC News
Our round-up of the latest news in high-performance computing
HPC Products 15
A brief update of the HPC software and hardware hitting the market
UK HPC centres 16
Stephen Mounsey on the UK’s most prominent HPC facilities
China grabs top spot in Top500 list 12
The 36th edition of the Top500 list of the world’s most powerful supercomputers has been published, with the top spot taken by the Chinese Tianhe- 1A system at the National Supercomputer Center in Tianjin, achieving a performance level of 2.57 petafl op/s (quadrillions of calculations per second).
The former number one system – the Cray XT5 Jaguar system at the US Department of Energy’s (DOE) Oak Ridge Leadership Computing Facility in Tennessee – drops to number two, achieving 1.75 petafl op/s running Linpack, the Top500 benchmark application. Third place is now held by a
Up to speed 20
Paul Schreier keeps up with the latest GPU developments
SGI has announced that the USA’s Los Alamos National Laboratory (LANL) has selected SGI to provide a solution to enhance an existing unclassifi ed institution-wide computing capacity. This solution features two large SGI Altix XE 1300 clusters.
Cover montage: Dan Taylor News in brief
l Isilon has announced that the Canadian Light Source, Canada’s national synchrotron research facility, has deployed one of the company’s scale- out storage systems to power
its research l Two US government agencies have become the fi rst customers to purchase Biomatters’ newly-launched sequence analysis software
package, Geneious Server l A new high-performance supercomputer has been installed at the University of Kentucky, putting the school among the USA’s top 10 public universities for computing power
12 The LANL Institutional Computing
project, the Climate, Ocean, and Sea Ice Modelling Project (DOE Offi ce of Science) and the Advanced Simulation and Computing programme (National Nuclear Security Agency) selected the SGI Altix XE clusters to
Chinese system called Nebulae, which was second in the June 2010 Top500 list. Located at the National Supercomputing Centre in Shenzhen, Nebulae performed at 1.27 petafl op/s. Tsubame 2.0 at the Tokyo Institute of Technology is at number four, having achieved a performance of 1.19 petafl op/s. Tsubame is the only Japanese machine in the top 10. At number fi ve is Hopper, a Cray XE6 system at DOE’s National Energy Research Scientifi c Computing (NERSC) Center in California. Hopper just broke the petafl op/s barrier with 1.05 petafl op/s, making it the second most powerful system in the US and only the
increase LANL’s HPC computing resources in the laboratory’s open computing environment. Each of the two Altix XE clusters
provides about 50 Tfl ops and 14 terabytes of memory with more than 4,500 processor cores. The fi rst cluster is already delivered and is being integrated into LANL’s existing Turquoise network infrastructure that supports institution-wide unclassifi ed scientifi c research and the ASC Predictive Science Academic Alliance programme. Built with Intel Xeon 5500 processors, Altix XE offers high
third US machine to achive petafl op/s performance. Of the top 10 systems, seven achieved performance at or above 1 petafl op/s. Five of the systems in the Top 10 are new to the list. Of the top 10, fi ve are in the US, with the others in China, Japan, France, and Germany.
WEB-EXCLUSIVE NEWS FEATURES AND REVIEWS NOW ONLINE
l Venus-C, a pan-European programme designed to integrate cloud computing with European Research Grids, is about to announce an open call for project bids.
www.hpcprojects.com Los Alamos National Laboratory selects Altix XE 1300s
performance, extensive memory and enhanced storage capacity. ‘LANL’s research requires top- notch computing and infrastructure for processing massive amounts of data,’ said Andy White, deputy associate director of the Theory, Simulation and Computing Directorate at LANL. ‘We have had excellent
experience in the past with SGI systems and are pleased to be deploying systems and working with them again to provide the computing resources required to support our important research initiatives.’
New EU headquarters for Promise Technology
Storage and RAID specialist Promise Technology has moved to new European headquarters in The Netherlands. The new headquarters doubles the space of the previous building, providing customers throughout Europe with a greater range of facilities as well as giving Promise Technology a foundation for future expansion. In particular, a new laboratory will house extensive test
SCIENTIFIC COMPUTING WORLD DECEMBER 2010/JANUARY 2011
environments for storage and RAID products for vertical markets, such as video monitoring, high-performance computing and video editing – in addition to demonstrating Promise Technology’s established server virtualisation, archiving and backup systems. Excellent logistical and transport links will provide fast replenishment of spares to customers throughout the
EMEA region. A training suite will allow Promise Technology to host a series of advanced technical courses, ensuring system integrators; system houses and dealers fully understand the latest storage technologies and how to implement them. Promise Technology will continue to develop its three established branch offi ces, in Germany, Italy and the United Kingdom.
www.scientifi
c-computing.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36