Today's supercomputers are built commodity-of-the-shelf (COTS) systems, are they not? Absolutely not, as the example of the new hpcLine system in Paderborn shows. Although the processors and many other components are standard, and can be found in many other computers too, the complete system is custom designed. This is typical of today's cluster supercomputers. The hpcLine in Paderborn was designed by PC2 - the Paderborn Parallel Computing Centre in collaboration with Fujitsu-Siemens Computers. Fujitsu Siemens led a consortium with many other companies that each delivered a part of the system. The hpcLine was inaugurated on June 21st at an event in Paderborn that also included presentations from supercomputing experts from Japan and the US. About one hundred attendees came to the city on one of the hottest days in Germany in June to learn about the cooling especially designed for the Paderborn hpcLine. The next day, at the presentation of the new official TOP500 list, we learned the system entered the list at position 205. Read further...

The new hpcLine supercomputer in Paderborn will be part of the high-performance computing equipment in the German state of North Rhein Westphalia (NRW). Paderborn is located in the North of that state. From the western part, Aachen, Christian Bischof, Director of the Supercomputer Centre at the Aachen University of Technology, explained at the hpcLine opening seminar, the ICT cooperation that exists between the academic institutions in the state in what is called the Ressourcenverbund NRW. Read further...

At the inauguration in Paderborn of the new hpcLine cluster, PC2 manager Jens Simon explained the design choices made in the cluster and showed some benchmark results. With 1,978 Tflop/s of Linpack performance, the machine ranks at position 205 of the TOP500 list of the world's most powerful supercomputers. Read further...

The Paderborn Center for Parallel Computing has a long history of providing services and doing research on HPC systems and Grid computing. The combination of the two is what makes the centre special. It is not a classical supercomputing centre, and it is not a tradional research centre, but it combines the two. Hence the systems, like the new hpcLine, are both research machines, exploring new architectures, and providing the basis to do research on new parallel or resource management software. Odej Kao from PC2 presented some new research projects that fit the tradition. Read further...

At the inauguration meeting of the hpcLine in Paderborn, Motoi Okudo from the Peta-Scale Computing Research Center at Fujitsu Laboratories in Japan, gave an overview of the Grid activities in Japan and explained the route the country is taking towards petaflops (Pflop/s) supercomputers. Current supercomputers are a thousand times slower: in the Tflop/s range. Japan has not given up the supercomputer race, and the approach is different from the US route. Read further...

Astronomy is one of the most ancient sciences that humanity has ever pursued. Since the dawn of civilisation, mankind has gazed at the night sky, awed and inspired by its black emptiness, the apparent immutability of the heavens, and wondered. Through the study of the heavens mankind has always sought to discover its place in the scheme of things, its connection to the divine, and its ultimate fate. Mankind has evolved from these early times, we live now in a technological era of high precision instrumentation, of large scale infrastructures to pursue knowledge. Yet we still gaze at the skies, wondering where we come from, and where ultimately we will go. The goal of modern Astronomy is still to answer these ancient, meaningful questions. It is a quest for knowledge that is driven by mankind's endless curiosity, by its need to understand, and it is fuelled by its inventiveness and ingenuity. New technological advances lead to new discoveries, but also to new questions.

Virtual research collaboration is being taken to a new level with the creation of the very first international, interdisciplinary computer Grid. It has been established by the Worldwide Universities Network - at five sites in three countries more than 5000 miles apart. This major step forward for Grid computing shows that real-world applications can be delivered on an international basis.

A new patch enhances TORQUE Resource Manager. TORQUE Resource Manager's user community, in co-operation with Cluster Resources, contributed several fixes in the last month to create TORQUE 1.2.0p5. One important fix stops the problem of qstat hanging while pbs_server is waiting for an update from an offline MOM. The problem occurred when a MOM in the "down, job-exclusive" state was marked offline using the "pbsnodes -o" command.

General Dynamics C4 Systems, a business unit of General Dynamics, and Lockheed Martin have held the U.S. Army's preliminary design review (PDR) on the Warfighter Information Network-Tactical (WIN-T) programme, the Army's next-generation battlefield communications network. During the PDR, the Army reviewed the "first draft" of the WIN-T system design, including the top-level architecture and key component technologies.

SANRAD Incorporated and emBoot have launched a joint solution that allows storage administrators to use SANRAD's iSCSI V-Switch in conjunction with emBoot's netBoot/i software to enable servers to boot off the IP SAN using any standard Ethernet network interface, eliminating the need for internal server disk drives. The SANRAD and emBoot solution extends Windows-based network booting using iSCSI into a wide range of storage, including Fibre Channel, FC SAN and SCSI.