The Japan Atomic Energy Research Institute (JAERI) will introduce an SGI Altix supercomputer as its new core computation system to assist in developing leading energy systems. As a result of a competitive bidding process, Fujitsu Limited in co-operation with SGI Japan will deliver to JAERI the new SGI Altix 3700 Bx2 model which is based on 2,048 Intel Itanium 2 processors, the Linux operating environment and over 13 terabytes of memory - the world's largest memory capacity. The system is scheduled to fully operate at the end of March 2005.

PathScale has introduced the PathScale OptiPath MPI Acceleration Tools. OptiPath incorporates the best MPI programming expertise available, and gives scientists, researchers and MPI experts direct visibility into complex MPl programme and cluster behaviour. It automates performance analysis and provides specific expert-level tuning advice that can be used to improve time to results and cluster efficiency.

Microway, a major HPC vendor specializing in 64-bit Linux clusters, servers, and RAID storage solutions, has partnered with PathScale Inc. to resell the PathScale InfiniPath Interconnect. Microway will incorporate PathScale's InfiniPath in their industry-leading four and eight processor Navion offerings. The Navion cluster platform, based on AMD's Opteron processor, was named "Best 64-bit Turnkey Solution" at ClusterWorld 2003.

Nexsan Technologies exhibited its award-winning, high-performance SATAblade and ATABeast disk-based products at

At the Electronica 2004 trade show in Munich, SBS Technologies Inc. continued its role as a single board computer market leader by introducing the CL9 high performance single board computer (SBC). The CL9 is a powerful processor platform for a wide range of applications and markets including multimedia, automation, transportation, imaging, medical, robotics and many others.

OptimaNumerics, specialized in innovative high performance technical and scientific computing software, has released OptimaNumerics Libraries 3.0, providing customers with new improved high efficiency numerical tools to maximize performance and make high performance more affordable to a broader range of scientific and technical computing users. OptimaNumerics is a Qubis Ltd. backed company.

Bull NovaScale servers have set new I/O performance records with Lustre, as a result of its strong co-operation with DataDirect Networks, Quadrics and Cluster File Systems. Bull here demonstrates its ability to deliver powerful solutions for high-performance computing.

The SC|05 High Performance Computing (HPC) Analytics Committee has announced the start of a year-long initiative to showcase more visibly the growing area of High Performance Computing Analytics within the annual Supercomputing conference.

Netherlands-based SARA Computing and Networking Services have successfully deployed Topspin's InfiniBand-based Server Switches as the interconnect fabric for its new 275-node server cluster, located in Almere, The Netherlands. SARA Computing and Networking Services, the Dutch National Supercomputing Center, has supplied for more than 30 years a complete package of high-performance computing, high-performance networking and infrastructure services throughout The Netherlands and Europe. Among SARA's customers are scientific, educational, and government institutions and the business community.

The National Institute of Nuclear Physics (INFN) is Italy's first disclosed customer for the new Cray XD1 supercomputer. Financial details were not disclosed.

Cray has received an order for a Cray XD1 supercomputer from Germany's Zuse Institute Berlin (ZIB). ZIB will use the Cray XD1 system primarily for computational chemistry applications. ZIB will also exploit the system's unique field programmable gate array (FPGA) capabilities to advance the use of reconfigurable computing to accelerate life sciences applications.

Concluding a 15-week effort with NASA and Intel to build and successfully install the world's most powerful supercomputer, the new 10,240-processor Columbia supercomputer is fully deployed at the NASA Advanced Supercomputing (NAS) facility located at Ames Research Center in Mountain View, California.

Streamline Computing Ltd, a UK provider of high performance cluster supercomputers will develop and market its software under the Allinea brand.

ClusterVision, specialist in Linux supercomputer clusters, has received an award from the Netherlands British Chamber of Commerce in the 23rd Anglo-Dutch Awards for Enterprise. This prestigious award, in the category for "Dutch Exporters" was presented by Heineken CEO, Mr. Anthony Ruys, Thursday November 18, 2004, at the Institute of Directors in London.

An expert in Semantic Web-based technology for enterprise resource interoperability (ERI) solutions, Metatomix Inc.'s mtx ERI PLATFORM is one of the first fully featured, node-based, architectures designed to take full advantage of Sun Microsystems' new Solaris 10 Operating System. The solution will enable "Computing on the Edge", the ability to utilize a distributed, heterogeneous, environment for real-time Enterprise-wide solutions.

The Cornell Theory Center (CTC) is demonstrating its latest research and development, software, and technologies at SC04, held at the David L. Lawrence Convention Center in Pittsburgh, from November 8-11. This marks the conference's 16th year and CTC's 15th year of participation.

The newly announced Solaris 10 Operating System from Sun Microsystems is PRIMEPOWER ready. The PRIMEPOWER-Solaris combination ensures customers that their Business Critical Computing environments, running mission-critical applications and databases – stay up and running securely, twenty-four hours a day, 365 days a year. Sun plans to make Solaris 10 available in January 2005.

Sun Microsystems in 2004 rolled out the most innovative systems portfolio in the company's 20-plus year history, the company said.

HP and Brocade Communications Systems have introduced an integrated offering designed to reduce acquisition and operating costs in industry-standard blade environments for enterprise and small and mid-size business customers.

Fujitsu Asia will launch CentricStor, the world’s first self-contained virtual tape appliance. The solution is designed to support major system platforms, tape automation platforms and tape devices in heterogeneous IT environments. Enterprises requiring high performance data back-up and restoration capabilities and considering introducing a disk-to-disk-to-tape backup system will find an unrivaled solution in CentricStor.

According to Bob Hertzberger, leader of the Dutch Grid project VL-E, Virtual Lab, the Grid and Web services communities will not come together. In fact, he saw at the IST conference in The Hague last week, clear indications that the Semantic Grid and Semantic Web Services communities are already drifting apart. Hertzberger told this at the European Grid Summit that takes place in The Hague, this week.

Lofar is a combination of sensors, fast network (Géant) and Blue Gene supercomputer processing power. Lofar is a long term project. The first phase starts now in The Netherlands. For the second phase, the Lofar consortium is in the process of acquiring funding.

The e-Infrastructures Reflection Group (e-IRG) could play an important role in shaping the future of Grids for eScience in Europe. At the European Grid Summit in The Hague, Ulf Dahlsten from the European Commission, proposed that e-IRG could produce and maintain a road-map for Grids and e-infrastructures, that could be used as input for the next Framework 7 programme of the European Commission and perhaps influence the current programme too.

The Apache project, well known for its web server and many Java and XML based open source reference implementations, is considering to develop reference implementations for the WSRF, WS-Notifications, and WSDM Grid standards currently under development. This was announced by Mark Linesh at the European Grid Summit that takes place this week in The Hague, The Netherlands.

Today the German communication and cluster management software company ParTec with its ParaStation starts to form a ParaStation Consortium. The partners will be Research Centre Jülich, University Wuppertal, University Karlsruhe and Partec with its new name "Cluster Competence Center" (CCC). Additionally they will offer ParaStation as Open Source based on the Sun Community License. The partners want to broaden the spectrum and the power of ParaStation, a cluster communication and management software.

Spacecraft constantly scan the Earth, creating hundreds of gigabytes of new data products daily. Working with this ever-growing mass of information has made ESA's Earth Observation Directorate a pioneer user of powerful Grid computing. Now ESA is participating in an ambitious project that applies this same technique to information distribution and retrieval, with the aim of creating 'Digital Libraries' for global scientific collaboration.

Few large production Grids are in place in the world. Even less have a large user population consisting of several types of users and seldom they ask the users in a survey what their experience is with the Grid. The Swedish Grid (SweGrid) is an exception. Anders Ynnerman, Director of the Swedish National Infrastructure for Computing, at the SEE-Grid policy workshop during the Europan Grid Summit in The Hague, The Netherlands, presented results from a user survey of this production Grid.

The Rocks 3.3.0 has just been released and supports all the favourite CPUs including athlon, itanium, nacona (em64t), opteron, and pentium. Everything is CDROM based, except for itanium which has a DVD for its base.

CSC, the Finnish IT Centre for Science takes part in one of the most important moves to bring together national supercomputing infrastructures to advance science and technology in Europe. Several leading European HPC centres have devised an innovative strategy to build a terascale supercomputing facility with continental scope, called Distributed European Infrastructure for Supercomputing Applications (DEISA). The resulting system will consist of more than 4000 processors, a huge memory space and an aggregate computing power of over 22 teraflops.

BEgrid, the infrastructure resulting from the Belnet Grid Initiative, has evolved to a pilot Grid environment on which the first applications start to run. Belnet and the BEgrid participants want to share the BEgrid experience with the whole research community and beyond. Hence a BEgrid seminar is organised on the 20th of December, 2004 at the Catholic University of Leuven (KULeuven) in Belgium.

The CAHPC showcased a number of research projects related to high performance computing and Cyber-infrastructure. But MammoData, a web accessible Grid tool for collaborative breast cancer diagnosis, made an immediate impression on spectators, due in part, to a flashy web-accessible user interface.

GridIron Software will work with IBM to bring the GridIron XLR8 Grid computing solution to enterprises for applications such as digital content creation, business intelligence and electronic design automation. GridIron XLR8 is an application development tool and runtime software that makes it simple to develop, use and manage software with the added speed of parallel distributed computing using IBM eServer, pSeries and xSeries servers running Linux.

At long last, the 4.0 beta 15 version has been released. The Cactus team expects to release 4.0 beta 16 in December, and a release candidate in January 2005.

Voltaire introduced a new InfiniBand switching solution. The Voltaire ISR 9096 features 96 InfiniBand ports of 10 Gbps (4X) or 32 ports of 30 Gbps (12X) full bisectional bandwidth in a modular chassis enabling high performance clusters and Grids.

The Optical Internetworking Forum (OIF) ends 2004 with a series of new work items at the highly successful and well attended fourth quarter meeting in October. The OIF's Architecture and Signaling working group launched a new work effort to develop a design guide for interworking control planes between network domains specified by OIF/ITU ASON and IETF GMPLS. The Physical and Link Layer (PLL) working group started a new work project on multi-level modulation (MLM) technology. On the opening day of the meeting, the OIF hosted the first in an ongoing series of workshops designed to advance information exchange and co-operation between the OIF and other industry groups. This initial workshop featured the Global Grid Forum (GGF).

Massachusetts General Hospital (MGH), a teaching hospital of Harvard Medical School, and IBM are working together to study how the development of a Grid-based, distributed computing infrastructure can facilitate improved collaboration and information sharing among cancer researchers.

Callidus Software's TrueComp Suite has successfully completed the IBM Grid enablement validation.

HP and Lockheed Martin have signed a strategic global alliance to drive new business opportunities for both companies around the world. HP and Lockheed Martin share a long history of co-operation on major U.S. Department of Defense (DoD), intelligence and civilian agency programmes. In the past three years alone, the companies have partnered on more than $500 million worth of business, including programmes in the United States such as the FBI's national Integrated Automated Fingerprint Identification System, the DoD's Defense Civilian Personnel Data System, and the Postal Service's Integrated Data System.

This article reports on some real results on the Earth Simulator and the early results from the IBM Blue Gene/L in the meteorology field. The low power consumption and small space of the Blue Gene/L (typical of T-type systems) are impressive, but the recent Linpack results and $/Teraflop/s metrics are likely to be overoptimistic as far as capability and productivity is concerned. The small memory on the node (4MB) will severely restrict efficiency on a large range of applications, not embarrassingly parallel. It raises the question as to whether the full configuration of 65,536 processors (360Tflop/s peak) Blue Gene/L system would come anywhere near achieving productivity (sustained performance) as good as the (40Tflop/s) Earth Simulator, on complex meteorology models ... (Chris Lazou)

Bull is broadening its security and e-transactions solutions offer, with the launch of a new generation of cryptographic hardware and solutions for Public Key Infrastructures. At the heart of this new offer, a USB security key will be the first highly secure individual and universal confidence tool.

Xilinx Inc. has made available two more Virtex-4 devices from the world's first multi-platform FPGA family. The Virtex-4 LX60 is the second device from the logic-optimized LX platform, and the SX35 is the first device from the DSP-optimized SX platform.

SciComp Inc. has been issued U.S. patent number 6,772,136 for its SciFinance software synthesis technology for financial instrument modelling using Monte Carlo simulations. This patent is in addition to a patent awarded to SciComp for the same problem solving environment, but using systems of partial differential equations.

Screaming past Japan's venerable Earth Simulator and 498 of the world's 500 most powerful systems, NASA's 10,240-processor Columbia supercomputer, built from 20 SGI Altix systems, seized the second position in the latest roster of the world's Top 500 supercomputers. Announced at the Supercomputing Conference 2004, the 24th Top 500 list confirms that NASA's Columbia supercomputer has added yet another performance milestone for the acclaimed SGI Altix platform from Silicon Graphics. Columbia is the world's fastest system to be based on industry standard Intel Itanium 2 processors and the Linux operating system. Currently, there are more than 800 SGI Altix systems delivered and in production worldwide.

After achieving international honours and accolades for building System X, the fastest supercomputer at any academic institution in the world according to the November, 2003 TOP500 List, Virginia Tech's rebuilt System X is now operating at 12.25 Tflop/s.

The US National Nuclear Security Administration (NNSA) Deputy Administrator for Defense Programmes Everet Beckner announced three supercomputers at its weapons labs are the among the six fastest computers in the world. Citing the recently released TOP500 supercomputer list, Beckner pointed out that NNSA labs not only topped the list with the BlueGene/L, but also claimed fifth with Thunder at Lawrence Livermore National Lab, and sixth with ASCI Q at Los Alamos National Lab.

PathScale has introduced the PathScale InfiniPath Interconnect, the industry's lowest latency Linux cluster interconnect delivering SMP-class performance for commodity-priced clustered computing. The PathScale InfiniPath Interconnect leverages three important industry standards, HyperTransport, InfiniBand, and the AMD64 architecture, to maximize performance and make low-latency interconnects more affordable to a broader range of high-performance computing users.

ProCurve Networking by HP has introduced a line of stackable switching products designed to enable enterprises to securely deploy the ProCurve Adaptive EDGE Architecture via Gigabit Ethernet to the network edge. The switches also feature new 10-Gigabit switch modules and accessories, which combine to minimize network delay and enable users to take full advantage of desktop and notebook capacity.

SURFnet and Internet2 have deployed industry-leading optical solutions from Nortel to provide optical interconnect capabilities into their internationally acclaimed research and education networks. The combination of these two deployments will allow for unprecedented secure and reliable collaboration among researchers and scientists globally.

Force10 Networks' TeraScale E-Series powered the winning demonstrations at SuperComputing 2004's annual Bandwidth Challenge, including the demonstration that set a new land speed record.

S2io Inc.'s Xframe product played a key role in setting a new record for sustained performance in the recently concluded Bandwidth Challenge held at this year's Supercomputing Conference in Pittsburgh.