| News digest April 2006 |
|
|
 |
| |
| Industry
-
HPCN industry |
|
|
| India's Institute for Plasma Research will use a Cray X1E Supercomputer for Theoretical Physics Studies |
|
India's Institute for Plasma Research (IPR) will acquire a Cray X1E supercomputer for conducting studies that include basic research in plasma physics and contributions to practical applications. Plasma physics, or the study of electrically conductive ionized gas, is the focus of research at the institute that supports studies of magnetically confined plasmas, material surface modifications and various theoretical modelling. The Cray X1E system was ordered through Cray partner Hinditron Infosystems, part of Hinditron Group.
Read further... |
|
|
|
| Fastest supercomputer in Japan installed at KEK - combination of Hitachi and IBM at 60 tflop/s peak |
|
The Japanese High Energy Accelerator Research Organization (KEK) is promoting simulation studies of high energy accelerator science including elementary particle physics and nuclear physics. In order to further accelerate the research, KEK has installed a new supercomputer system and starts its operation on March 1st, 2006.
Read further... |
|
|
|
| ClearSpeed starts product shipments for Japan's near-future largest 85 Tflop/s supercomputer at Tokyo Institute of Technology |
|
ClearSpeed Technology announced the company is delivering the first shipment of ClearSpeed Advance boards to Sun Microsystems to be included in Japan’s largest supercomputer. The cluster is being built for the Tokyo Institute of Technology (Tokyo Tech), is anticipated to be one of the ten largest supercomputers in the world at 85 TeraFLOPS, and is expected to be ranked by the TOP500. ClearSpeed will be contributing 360 Advance boards for the system. Other vendors participating in this system build include AMD, NEC, Sun, and Voltaire. The system is expected to be up and running in Q2 of 2006.
Read further... |
|
|
|
| Green light for Europe's fastest supercomputer in Juelich |
|
A new era of supercomputers was inaugurated when Thomas Rachel, BMBF State Secretary, together with Prof. Andreas Pinkwart, NRW's Innovation Minister, gave the green light to the fastest supercomputer in Europe, "JUBL" (Jülich Blue Gene/L), with a power of 46 teraflops (trillion computing operations per second). Amongst the purely science-based research institutions, Research Centre Jülich is in fact number one in the world. The JUBL technology developed by IBM will open the door for completely new applications in biology, chemistry, physics and climate research to benefit material scientists, nanotechnologists and energy researchers.
Read further... |
|
|
|
| AWE's workload benchmark picks 40Tflop/s Cray XT3 |
|
Since 1950 the Atomic Weapons Establishment (AWE) has been central to the defence of the United Kingdom - providing and maintaining the warheads for the country's nuclear deterrent. Uniquely among the nuclear powers, AWE covers the whole lifecycle of nuclear warheads, in a single organisation. This includes initial concept, research and design, through component manufacture and assembly, to in-service support and, finally, decommissioning and disposal. (Chris Lazou)
Read further... |
|
|
|
| Cray's Adaptive Supercomputing; A Paradigm Shift |
|
"Adaptive supercomputing will cause a paradigm shift in the way users select and use HPC systems. Adaptive supercomputing is necessary to support the future needs of HPC users as their need for higher performance on more complex applications outpaces Moore's Law. The Cray motto is: adapt the system to the application - not the application to the system", says Steve Scott CTO of Cray Inc, March 2006.Cray Inc. are currently giving briefings on their vision of 'Adaptive Supercomputing' (See Cray Press Release 03-20-2006). I was briefed by Steve Scott so let me share with you, in broad terms, what he said. The increasing demand, for better performance, can no longer be achieved through Moore's law processor improvements and a one-size-fits-all system mentality. HPC users are no longer getting the performance advances they need from microprocessors. Commercial response to Moore's law slowdown has been to provide multi-core chips. These are general-purpose architectures, optimized for most widely used applications. But as it is widely recognized when scientific computing migrated to commodity platforms, interconnect speed, both in terms of bandwidth and latency, became the limiting factor on application performance and remains a bottleneck to this day. (Chris Lazou)
Read further... |
|
|
|
| DEISA high performance computing |
|
The DEISA (Distributed European Infrastructure for supercomputing applications) Consortium will hold a symposium on high performance computing in Bologna, Italy, on 4 and 5 May, 2006. The purpose of the event is to provide a forum for the discussion of scientific and strategic challenges in the field of high performance computing, and to assess the impact of DEISA on computational science in Europe.
Read further... |
|
|
|
| Quantum computer solves problem, without running |
|
By combining quantum computation and quantum interrogation, scientists at the University of Illinois at Urbana-Champaign have found an exotic way of determining an answer to an algorithm - without ever running the algorithm. Using an optical-based quantum computer, a research team led by physicist Paul Kwiat has presented the first demonstration of "counterfactual computation", inferring information about an answer, even though the computer did not run. The researchers report their work in the February 23 issue of the journalNature.
Read further... |
|
|
|
| Second generation of Bull NovaScale systems |
|
Bull announced the NovaScale 5005 series: the second generation of large NovaScale servers.The NovaScale 5005 servers are SMP (Symmetrical Multi-Processing) systems, built using modular blocks that can rapidly be combined within a single cabinet to provide SMP servers from 2 to 32 sockets.
Read further... |
|
|
|
| IBM achieves performance breakthrough in massive computer networks |
|
IBM has scored a performance breakthrough in a project code-named "Project Fastball" on the ASC Purple supercomputer - the third most powerful supercomputer in the world. ASC Purple, located at Lawrence Livermore National Laboratory (LLNL), is a National Nuclear Security Administration computing system designed by IBM. IBM and LLNL demonstrated over 102 gigabytes per second of sustained read and write performance to a single file using specialized software that manages the transfer of information between thousands of processors and thousands of disk storage devices. The world record performance was achieved using 416 individual storage controllers combined with 104 Power-based eServer p575 nodes.
Read further... |
|
|
|
| NEC develops vertical-cavity surface-emitting laser boasting world's fastest operation speed of 25 Gbps/Ch |
|
NEC Corporation has successfully developed a vertical-cavity surface-emitting laser (VCSEL) that boasts the world's fastest 25 gigabit per second (Gbps) per channel (ch) operation speed. It subsequently achieves ultra-high-speed optical interconnection among LSI chips/boards, representing a major step toward the realization of next-generation supercomputing systems.
Read further... |
|
|
|
| Swedish National Supercomputing Center to tackle large-scale problems with new SGI Altix system |
|
Swedish researchers have installed the country's largest shared-memory supercomputer, a new 64-processor SGI Altix system equipped with half a Terabyte of memory. Deployed this month at the National Supercomputing Center (NSC) at Linkoping University (LiU), the new supercomputer will allow physicists and other researchers from throughout Sweden to break through computational barriers created by complex computations.
Read further... |
|
|
|
| ORNL researchers and supercomputer earn big roles in new DOE projects |
|
More than 2 million hours of processing time on some of the world's most powerful computers could help Oak Ridge National Laboratory researchers make significant strides in astrophysics, fusion and materials science. Two of the three ORNL projects funded through the Department of Energy's recently announced Innovative and Novel Computational Impact on Theory and Experiment (INCITE) programme involve collaborations with universities, industry and other national labs. The third belongs solely to ORNL and is led by Phani Nukala of the Computer Science and Mathematics Division.
Read further... |
|
|
|
| New Cray X1E Supercomputer at India's NCMRWF will be the most powerful weather prediction system in the region |
|
India's National Centre for Medium Range Weather Forecasting (NCMRWF) will acquire a Cray X1E supercomputer that will rank as the subcontinent's largest and most powerful numerical weather prediction (NWP) system. NCMRWF, an agency within India's Department of Science and Technology located in Noida near New Delhi, recently ordered the system through Cray partner Hinditron Infosystems, part of Hinditron Group.
Read further... |
|
|
|
| Air Force Research Labs to boost 'Eyes in the Sky' with Star-P |
|
The Air Force Research Laboratory (AFRL) in Rome, New York, is developing a new supercomputer-based system that could help the military analyse massive amounts of satellite-based radar data for surveillance, missile warning, communications and navigational activities.
Read further... |
|
|
|
| Cray announces adaptive computing strategy |
|
Cray announced plans to develop supercomputers that will take the concept of heterogeneous computing to an entirely new level by integrating a range of processing technologies in a single platform. These "adaptive supercomputing" systems will be able to solve scientific and engineering problems more quickly - and make programmers and end users more productive - by adapting processing to the requirements of each application.
Read further... |
|
|
|
| Advance registration opens for 21st International Supercomputing Conference in Germany, Conference Programme now available on-line |
|
Advance registration - at lower rates - is now open for the 21st International Supercomputer Conference (ISC2006) to be held June 27-30 In Dresden. Under the theme of "Addressing Challenges for Industry, Academia and Research", ISC2006 offers three days of intensive technical sessions and plenty of networking opportunities to provide attendees with a comprehensive and compact overview on HPC topics presented by world-renowned scientists and experts. New this year will be panel discussions of key topics and birds-of-a-feather sessions where attendees can gather to discuss areas of special interest. On Tuesday, June 27, three tutorial programmes will be presented.
Read further... |
|
|
|
| SC06 sends out the 2006 Call for Participation |
|
SC06, the annual conference for high performance computing, networking, storage and analysis has sent out a call for participation. The SC06 conference will be held November 11th -17th , 2006, at the Tampa Convention Center in Florida. The deadline for submissions in these areas is Monday, April 17.
Read further... |
|
|
|
| NASA awards supercomputing time for advanced research |
|
NASA has awarded approximately 4.65 million hours of supercomputing time to researchers. The computing time may help scientists solve some of the most challenging research problems involving climate variability, combustion burners, flow conditions and novel electronic materials.
Read further... |
|
|
|
| Comcast deploys CallMiner's Virtual Server Room to drive customer satisfaction initiatives |
|
Comcast Corporation, a provider of cable, internet, and IP-enabled phone service, is deploying CallMiner's Virtual Server Room (VSR) distributed processing platform to analyse 100% of recorded calls in their Midwest Call Center. CallMiner's VSR makes possible medium and large speech analytics installations, which were once considered too cost and resource-intensive to deploy.
Read further... |
|
|
|
| TORQUE 2.0.0p7 increases usability and reliability |
|
Cluster Resources has released its latest version of TORQUE. TORQUE 2.0.0p7 includes improvements in scalability, usability, and reliability. In addition, the on-line TORQUE documentation has been updated and the new TORQUE WIKI is also available.
Read further... |
|
|
|
| CSC to provide technical and management expertise for High Performance Computing in Ohio |
|
AE-PR-04-06-20Computer Sciences Corporation announced today that it has been selected by Advanced Virtual Engine Test Cell (AVETEC) to serve as the management and integration contractor on the Data Intensive Computing Environment (DICE) effort. Under the terms of the agreement, CSC will provide management, integration, authentication systems, architecture, operations, project selection and communications outreach to assist the newly established DICE initiative.
Read further... |
|
|
|
| Sun shows world-record performance result for Sun Fire systems running UltraSPARC IV+ processors |
|
The Sun Fire E25K server running UltraSPARC IV+ processors and Solaris 10 Operating System achieved world record performance of over 1.16 million business operations per second on the SPEC Java Business benchmark (SPECjbb2005). The latest in a series of leading performance recordsS, testing results were conducted on the Java Platform, Standard Edition 5.0_06 with Java HotSpot Virtual Machine technology, the 14th world-record benchmark on the UltraSPARC IV+ processor-based platforms in just six months since launch.
Read further... |
|
|
|
| Linux Networx announces Linux supercomputing storage solutions |
|
Linux Networx has made available storage solutions optimized for a wide range of supercomputing needs. Integrating the appropriate storage configuration within a supercomputing system has been proven to power performance acceleration and utilization increases by a factor of 20 or more.
Read further... |
|
|
|
| Mitrionics' FPGA supercomputing platform now compatible with Xilinx Virtex-4 platform |
|
Mitrionics's Mitrion Platform, consisting of the Mitrion Virtual Processor and Mitrion Software Development Kit, now fully supports the Xilinx Virtex-4 Platform FPGAs. The Mitrion Platform, in conjunction with Virtex-4 FPGAs, allows both of these requirements to be met as supercomputing applications are able to be written by software developers without any hardware design knowledge. Supercomputing performance is generally 10x to 100x greater than traditional processors with applications targeted towards the industries of bioinformatics, oil and gas, imaging, and financial analysis.
Read further... |
|
|
|
| HPC User Forum invites European user talks |
|
European HPC users engaged in leading-edge activities are invited to contact the HPC User Forum to secure speaking slots at one of the next U.S.-Europe dialogue meetings. The meetings will be held on May 29-30, 2006 at CINECA in Bologna, Italy and June 1-2, 2006 at ETH in Zurich, Switzerland. All interested members of the HPC community in Europe are invited to attend one or both meetings without charge.
Read further... |
|
|
|
| Purdue University selects IBRIX Fusion to boost its supercomputing centre performance power |
|
Purdue University has deployed the IBRIX Fusion Parallel File System, an enterprise-class, software-based scalable file serving solution for its three primary clusters, totaling 914 nodes and nearly 14 teraflops, located at the University's Rosen Center for Advanced Computing.
Read further... |
|
|
|
| Verari Systems announces support for new dual-core AMD Opteron processors |
|
Verari Systems, a developer of blade computing systems, will increase the company’s support of AMD’s dual-core chip offerings by adopting the fifth set of dual-core processors released from AMD consisting of Dual-Core AMD Opteron processors Models 185, 285 and 885.
Read further... |
|
|
|
| COSCON transforms shipping business with SOA based on IBM software and services |
|
COSCON, China's largest container shipping company, is relying on IBM software and services to help streamline its business processes, increase efficiency and enhance collaboration with its customers, partners and suppliers.
Read further... |
|
|
|
| Pininfarina trusts Bull NovaScale Servers to host its HPC applications |
|
Pininfarina - historically a major Italian and worldwide automotive design company - has chosen Bull to supply a first tranche of NovaScale 4020 servers, based on Intel Itanium 2 processors, for the implementation of a high-performance computing (HPC) centre dedicated to managing the computer simulation of the project cycles/production processes for their own cars.
Read further... |
|
|
|
| Cray reports selected, preliminary fourth quarter and full-year 2005 financial highlights |
|
Cray's management, its audit committee and its auditors are reviewing an issue, identified by management, to determine if revenue recognized in 2004 under one of the company's product development contracts was appropriately recorded. As a consequence, the company has delayed filing its 2005 Form 10-K until this review is completed.
Read further... |
|
|
|
| Oracle and Fujitsu Siemens Computers achieve world record result on two-tier SAP BW benchmark on Fujitsu PRIMEPOWER |
|
Oracle and Fujitsu Siemens Computers anounced a world record result on the two-tier SAP Business Information Warehouse (SAP BW) Standard Application Benchmark.
Read further... |
|
|
|
| Mitrion first FPGA supercomputing platform to support new SGI RASC RC100 computation blade w/Dual Xilinx Virtex-4 FPGAs |
|
Mitrionics Inc.'s Mitrion Platform, consisting of the Mitrion Virtual Processor and Mitrion Software Development Kit, has been successfully tested and now supports the new SGI RASC RC100 computation blade, built with dual Xilinx Virtex-4 FPGAs.
Read further... |
|
|
|
| United Devices expands management team with distributed computing and HPC industry veterans |
|
Distributed computing veteran John Hime has joined United Devices' Board of Directors and high-performance computing (HPC) expert Mark Grefer has been named UD's senior vice president of sales.
Read further... |
|
|
|
| HP upgrades Integrity servers and speeds software deployment on HP-UX operating system |
|
HP has enhanced its HP Integrity server line and HP-UX 11i operating environment with significant capacity, virtualization and management upgrades. The enhancements offer customers greater server capacity as well as faster deployment of enterprise software within an HP Virtual Server Environment (VSE) running on the HP-UX 11i operating system - some virtualization projects can be brought on-line in less than half the time.
Read further... |
|
|
|
| |
| Industry
-
The Grid |
|
|
| French Grid'5000 will employ over 3000 processors this year |
|
The French Grid'5000, the country's national Grid effort currently deploys over 2000 processors in 17 laboratories at 9 sites in Bordeaux, Grenoble, Lille, Lyon, Nancy, Orsay, Rennes, Sophia-Antipolis, and Toulouse. Later this year, the number of installed processors will reach 3000 with over 2 Tbyte of memory, and 150 Tbyte of disk, Franck Cappelo told the audience at the Trends in High-Performance Distributed Computing workshop in Amsterdam. The ultimate goal is to have 5000 processors installed, hence the name Grid'5000. The effort is comparable to TeraGrid in the US, NaregiGrid in Japan, and EGEE at a European level. Each of these efforts takes a different approach. In France they choose to build the Grid as a nationwide computational cluster, isolated from the Internet. This has the advantage that the environment can be easier controlled and that one can concentrate on the many experiments that are done on Grid'5000. Experiments that range from computer science (networking, middleware and programming environments) to applications including computational electromagnetism and massively parallel computations in multi-material fluid mechanics.
Read further... |
|
|
|
| Japanese NAREGI Grid on schedule |
|
The Japanese Grid, called NAREGI, is a 100 million euro, 150 fte effort to create a nationwide standards based Grid. At the Trends in High-Performance Distributed Computing workshop in Amsterdam, NAREGI deputy director Satoshi Matsuoka explained the Beta 1 release of the Grid middleware is nearly completed. It is based on OGSA/WSRF/GT4 and implements also the GGF JSDL standard. NAREGI Grid concentrates on connecting supercomputing centres and national labs with the aim to solve grand challenge applications. NAREGI knows a strong industry participation. The NAREGI Grid will provide some 85 Tflop/s of processing power and 1.1 Pbytes of storage. Currently there are about 3000 CPU's with a combined power of 17 Tflop/s.
Read further... |
|
|
|
| Making public donation based Grids more reliable |
|
Public donation based Grids are getting more and more attention. Recently the BBC in the UK announced support for a climate research project. People can download a client from the BBC website and participate by donating computing time. However, there are also a number of people that try to manipulate the results that come from their computer: either because they want to get higher in the rankings, or just because of the "fun" of it. It is estimated that in for instance SETI@Home about 1% of the results returned by participating computers were wrong. The underlying Grid middleware, BOINC in these cases, cannot determine whether a result is correct or not. It uses some statistical waging after submitting a job to three different machines. A lot of capacity is lost in this way, and still one is not certain of the results produced. At the Trends in High-Performance Distributed Computing workshop in Amsterdam, Jon Weissman from the University of Minnesota presented the results of experiments in trust-sensitive scheduling in Open Grid environments. His statistical methods try to identify malicious nodes and exclude them from calculation on the fly. Even with 30% of the systems affected, the whole Grid system is back in order in a matter of seconds.
Read further... |
|
|
|
| Next wave of EU-funded IST Grid projects in negotiation phase |
|
The European Commission has announced that currently it is in negitation with twenty different consortia for funding their project proposal. The proposals were submitted in response to the FP6-IST call 5 in September 2005. The proposals address experimentation and adoption of enhanced Grid middleware by business and industry, Network-Centric Grid Operating Systems, agent- and semantic-based Grid architectures and middleware, knowledge Grids, Grid services and business models, trust and security, Grid developer platforms and user environments. There are also specific actions supporting the European Technology Platform NESSI in the area of Grid, Software and Service Oriented Architectures and the development of a Grid research vision and roadmaps. The contracts for these 20 proposals are expected to be signed this summer, with a formal launch event scheduled to take place in Brussels on 19 September 2006.
Read further... |
|
|
|
| Computing on demand in Europe |
|
The new Sun Grid computing on demand service attracts a lot of attention. You can run applications over the Internet. It is offered to US customers only. This type of service is not very new, however, companies like T-Systems and Hemeris offer that type of service for competing prices already for many years in Europe. So there is a choice here. In the US, there are also several companies provding comparable services, some at 60% of Sun's price.
Read further... |
|
|
|
| DDOS atttack on Sun Grid at first day of operations? |
|
At the first day of its operation, the Sun Grid was under attack by a denial of service attack that affected the demo application. In an answer to that, Sun has moved the application to the closed part of the Sun Grid. You have to sign up for that with a Paypal account to access the SUN Grid and the example application. Sun expects that will be sufficient to handle future DDOS attempts.
Read further... |
|
|
|
| HP extends Grid leadership | |
