Primeur is a monthly Virtual Magazine on High Performance
Computing
and
Networking in Europe. It is produced by an
editorial
team
composed of professionals in publishing and HPCN. You can put the editorial
team to
work as well. Read about our services and
advertising
opportunities
and find out about the friendly
rates
.
advertisement
Breaking news
- just a click away:
Choose a section or scroll down to the articles
|
|
|
 |
XML
|
|
|
The Grid
|
|
|
Media and visualisation
|
|
|
Linux
|
|
|
HPCN industry
|
|
|
Cluster computing
|
|
|
Applications
|
|
|
|
|
|
|
United Kingdom
|
|
|
Germany
|
|
|
Finland
|
|
|
|
|
|
|
| Smarr to leave NSCA
|
|
Larry Smarr has resigned as director of NSCA and of the US computing Alliance to take on a professorship at the University of California, San Diego. He will stay advisor of the Alliance for several activities. Smarr was the first to coin the term metacomputing and to talk about the Grid.
|
| Cray SV1 for
Alabama's scientific community
|
|
|
The
Alabama Supercomputing Authority will use its new 16-processor Cray SV1
supercomputer to simulate respiratory problems in infants, stock market behavior and other complex computer applications. The Authority recently bought the Cray system to keep pace with a rapidly growing demand for additional computing power and time from more than 150 user groups statewide.
|
| Shanghai supercomputer centre
|
|
|
Shanghai, China's leading industrial and commercial city, has begun to build a supercomputer centre as part of its efforts to turn itself into an international information centre The Shanghai Super Computing Centre, scheduled for completion by the end of the year, will be equipped with homemade supercomputers with
over 300 billion calculations per second. performance.
|
| Export of supercomputers from the USA made easier
|
|
|
In the USA, the House of Representatives
agreed to reduce from 180 days to 60 days the time Congress has to review administrative changes to export controls on high-speed computers.
|
| NEC develops
atomic holography technology that can manipulate
atomic images into animation-like movies
|
|
|
NEC researchers, together with Professor Fujio Shimizu of the University of Electro-Communications, have developed an "atomic beam movie," a new atomic holography technology that can manipulate atomic images in animation-like movies, for the first time in the world. NEC is looking to integrate the use of atom beam lasers with the results of the research to take further steps toward the development of production technologies enabling the use of both aluminum (Al) and germanium (Ge) circuitry in nanometer-level ultra LSIs and quantum devices of the future, that will avoid increased changes of mask patterns and steps required with conventional technology.
|
|
|
|
|
|
| Finnish supercomputer centre CSC buys Tflop/s IBM RS/6000 SP
|
|
|
The Finnish academic supercomputing center - CSC - has selected an IBM RS/6000 SP system as its next generation supercomputer for the Ministry of Education. Upon installation, the IBM system will be one of the most powerful supercomputers on the European continent. The machine will have a peak in the Tflop/s range in 2002, making it one of the fastest computers in Europe. The new IBM RS/6000 SP will be installed at the CSC in October 2000, replacing eventually a Cray T3E. WM-data, an IBM business partner, will install the system, valued at 30 million Finnish marks or 4.5 million euro.
|
| Compaq delivers 100 GFlop/s Alpha Cluster to CSC and Helsinki University of Technology
|
|
|
The Finnish Center for Scientific Computing (CSC) and the physics and technology laboratories of the Helsinki University of Technology (HUT) ordered a Compaq supercluster system that will be installed beginning of summer and forms one of the largest single hardware installation for scientific computing in Scandinavia. The supercluster system contains 96 nodes, Compaq AlphaServer DS10 servers with new generation Alpha processors. Additionally a DS20E (2 processors) serves as front end machine. As interconnect Myrinet and FastEthernet is used. The operating system is Compaq Tru64 UNIX, the peak performance greater than 100 GFlop/s.
|
|
|
| Bayer installs 1.26 Tbyte Shark storage server
|
|
|
Bayer,
one of the world's largest manufacturers of pharmaceuticals installed
a 1.26 Tbyte IBM Shark and consolidate the data from its IBM RS/6000 and AS/400 servers into a cost-effective solution enabling applications on both servers to run faster and make management easier.
|
| Mannheim Supercomputer Seminar for 15th time
|
|
|
The 15th Mannheim Supercomputer Conference will start
next week: June 8 - 10, 2000.
It is the
supercomputer event of Europe. Besides the main management oriented Seminar, there is also a tutorium in Grid computing and the RCI Aerospace/Automotive SIG Meeting an Aerospace and Engineering conference. Because the
'HPC- Centers and Manufacturers Forum' has had
an extremely positive resonance by the manufacturers, the organisation
decided this year to use this resonance as to the
arrangement of this year's 'Manufacturers Forum' as a
means of enabling conversation between manufacturers
and users. The event will be covered by a special PrimeurLive!.
|
| ViewTech and Dynamic Imaging merge
|
|
|
The two software companies, ViewTech and Dynamic Imaging have found each other.
The boards in the two companies have agreed upon a letter of intent, and the goal is to have a merger plan on the table within the end of May.
ViewTech will increase its core business, which is to further develop and sell the GLview Product Line, but the company will be operating under a different name and logo.
|
|
|
| UK high-end computing
|
|
|
The UKHEC Collaboration gathers and spreads information on international developments in computer software and programming tools. This includes programming languages (e.g. C++, Java, Fortran95), software engineering and software Quality Assurance, data management, visualisation and Virtual Reality, and metacomputing. Funding is provided by the Engineering and Physical Sciences Research Council (EPSRC)
|
|
|
|
|
| White Paper on High-End computing by Europan RCI advisor
|
|
|
RCI published a White Paper "High-End Computing in the 21st Century: A Personal View", written by Alessandro Formica, RCI European Scientific Advisor. This White Paper provides the results of more than two years of talks with researchers, academicians, and managers of European universities, research centers, and industry. The analysis of U.S. documents and programs and talks with U.S. researchers also played a key role in shaping this vision.
RCI has also released the final programme of their European Conference in June in Mannheim.
|
|
|
|
|
|
| Object Oriented finite element library Deal.II
|
|
|
The University of Heidelberg has released version 3.0 of the deal.II object-oriented finite element library. Deal.II is an object oriented finite element library implemented in C++. The library is aimed at research usage.
|
| Blue Mountain
achieves record-breaking run
|
|
|
Los Alamos' Blue Mountain supercomputer recently set a world record by running 17.8 years of equivalent single-processor computing in just 72 hours. The engineering calculations on the U.S. Department of Energy's supercomputer,
analyzed thousands of variables to simulate how well a nuclear weapon and its key components would survive upon impact with the ground. During the three-day period, more than 15,000 engineering simulations that required 10 hours each were executed across 31 of Blue Mountain's 48 SGI Origin 2000 servers, or 65% of the entire machine.
Each of the servers contains 128 processors.
|
| Travelling back to the beginning of time on a 696 processor supercomputer
|
|
|
An international team of cosmologists has released the first detailed images of the universe in its infancy.
The images reveal the structure that existed in the universe when it was a tiny fraction of its current age and 1,000 times smaller and hotter than it is today. In Europe, the University of Rome La Sapienza is involved in the project called
BOOMERANG. For the calculations,
NERSC's 696-node Cray T3E supercomputer was used.
Total computing time was less than three weeks on this machine.
|
| NEC shipped 50 SX-5 systems to date - new computer at University of Montreal
|
|
|
The Centre for Research on Computation and its Applications (CERCA) today announced that they began production use of a new NEC SX-5S/2 parallel vector supercomputer.
The University of Montreal acquired the system in support of RQCHP, a research consortium comprised of the University of Montreal, CERCA and the University of Sherbrooke. Research at CERCA is focused on aspects of fluid dynamics (aerodynamics, hydraulics, turbulence, environment and combustion) and chemistry (pharmaceutical chemistry and materials).
The computational resources are further augmented by a substantial visualization and networking infrastructure. Since first shipment in January 1999, 50 SX-5 systems have already been installed worldwide.
|
| PAM-FORM 2000 - 3D simulation of composite and plastic materials from ESI
|
|
|
ESI Group released PAM-FORM 2000, the first 3D software for forming simulation of composite and plastic parts. PAM-FORM enables virtual and realistic analysis and optimization of a variety of manufacturing processes, including thermoforming, dry-fabric stamping, draping, diaphragm-forming, blowing, extrusion-blowing and stretching-blowing. PAM-FORM simulates parts production using physical modeling based on CAD data, taking numerous critical parameters into account such as the materials used, friction, holding systems or even press speed or pressure cycles.
|
| RS/6000 fast on Baan
|
|
|
IBM
announced that its DB2 Universal Database running on IBM's latest RS/6000 servers has achieved record-breaking benchmark performance on a number of leading ISV solutions. In separate certified Baan tests, DB2 set record-breaking benchmark results on both RS/6000 F80 and S80 models. Lawson Software's enterprise e-business solutions showed extremely high data throughput and transaction performance running with DB2 Universal Database on the copper-based RS/6000 model M80 and S80. These results demonstrate that DB2 offers optimum scalable performance to customers.
|
| Comdisco installs 14 IBM S/390 servers
|
|
|
Comdisco has invested more than $30 million for IBM S/390 servers to support the availability and continuity requirements of its customer base. Comdisco is installing 14 new IBM S/390 G6 enterprise servers in facilities across North America, increasing its processing power by 450 percent.
|
| RS/6000 SP System
on top of TPC-H benchmark
|
|
|
IBM 's next generation RS/6000 SP system with DB2 Universal Database Version 7.1 has set a new record for business intelligence performance in the TPC-H benchmark, beating the previous record at a price performance ratio by a factor of three.
|
| Milinx next-generation ASP powered by Sun
|
|
|
Milinx Business Group, Inc, an Application Service Provider (ASP)
entered into an agreement with Sun Microsystems, Inc.
As part of the agreement, Milinx will also leverage Internet infrastructure software from
iPlanet E-Commerce Solutions, a Sun-Netscape Alliance, for its new next-generation ASP service package
|
|
|
| EWOMP 2000 - Second European Workshop on OpenMP
|
|
|
Submission deadline for the
Second European Workshop on OpenMP to be held
September 14-15th 2000, Edinburgh, Scotland,
is May 15. This workshop will be a forum for discussion of the latest developments in OpenMP and its applications. Topics of interest include:
OpenMP implementations;
Proposals for, and evaluation of, language extensions;
Applications development experiences;
Comparisons with other approaches such as MPI and HPF.
Benchmarking and performance studies;
Compilers, debuggers and performance analysis tools for OpenMP. Conference web site; http://www.epcc.ed.ac.uk/ewomp2000/
|
| General Micro Systems ships the fastest multiprocessor CPU board
|
|
|
General Micro Systems
are shipping the dual 850 MHz Atlas-C -- the first multiprocessor CompactPCI board based on the new Intel Coppermine-256 Flip Chip Processors. Atlas-C provides built-in support for symmetric and real-time asymmetric multiprocessing, which enables its two Coppermine-256 processors to work together in parallel on the same program, transparent to the programmer and the application.
|
|
|
| First Cray T3E order for new Cray Inc.
|
|
|
Cray Inc. ,
has been awarded an $18.5 million contract to upgrade the U.S. Army High Performance Computing Research Center's (AHPCRC) Cray T3E-1200 supercomputer. This marks the first major order for the company formed from the March 31, 2000, Tera Computer Company acquisition of Cray Research from SGI. The contract also includes an option for AHPCRC to acquire next-generation Cray SV2.With the addition of 816 processors to its existing 272-processor Cray T3E-1200
the AHPCRC will operate the largest Cray T3E system in the world. The expanded 1,088-processor system increases the computational capability of the AHPCRC's existing system to a peak performance of 1.3
Tflop/s and memory size of
557 Gbyte.
|
| SGI announces
management changes
|
|
|
SGI
announced two senior management changes. Warren Pratt has been named as senior vice president, engineering, responsible for product development and engineering. Betsy Rafael, senior vice president and chief financial officer, has resigned to pursue an opportunity with a private, non-competing firm. In the interim, SGI's senior financial team under the direction of Bob Bishop, chairman and chief executive officer, will manage financial and administrative operations.
|
| Sun and Veritas expand storage cooperation
|
|
|
Sun Microsystems
and
Veritas Software Corporation
announced that Sun will package Veritas NetBackup
software with
Sun StorEdge tape libraries
to deliver a comprehensive and fully automated backup management solution for data center environments.
|
| New GS top-of-the-line machines from Compaq
|
|
|
Compaq Computer Corporation
launched of its new high-end
AlphaServer GS series systems.
The GS series consists of high-end 8-,
16- and 32-way servers running the Tru64 UNIX and OpenVMS operating
systems. Targeting dot-coms as well as
customers in telecommunications, high-performance technical
computing, healthcare and financial services, the new
AlphaServers
are optimized to support the most demanding e-business applications
and customer environments. With over 200 systems already sold,
Compaq anticipates $1 billion
in revenue from
AlphaServer
GS
series sales during 2000 and expects to double its Unix market share
within target markets in the next three to four years.
|
| IBM claims world's fastest 6-way web werver
|
|
|
IBM
introduced what it claims to be
the world's fastest 6-way Web server - the RS/6000 H80 -- a rack-mounted dynamo that uses copper microprocessors to speed e-business on the Web.
In independent testing on the SPECWeb99 benchmark, a 6-way RS/6000 H80 running Zeus Web server software supported 2200 simultaneous connections.
|
| Last Tera results
|
|
|
Former Tera has announced the last financial results that do not include the Cray acquisition.
For the first quarter ended March 31, 2000, Cray Inc. reported a net loss of $8.0 million, on revenues of $43,000, compared to a net loss of $6.8 million,
on revenues of $661,000 for the first quarter of 1999.
|
| SGI to spin-off MIPS
|
|
|
After selling Cray to Tera, SGI now plans to spin-off all of its shares in processor producing MIPS. Currently SGI still holds 65% of the MIPS shares. The shares will be given as dividend to SGI's shareholders.
|
| Compaq workstations with
866 MHz processors
|
|
|
Compaq introduced new workstations with Intel Pentium III and Pentium III Xeon 866 MHz processors especially
for graphics and compute-intensive applications.
The new models are the Compaq Deskpro Workstation AP250 and Compaq Professional Workstations AP550 and SP750 .
|
| IBM debuts first servers powered by silicon-on-insulator technology
|
|
|
IBM today announced a new line of AS/400e servers powered by the world's first production microchips made of
silicon-on-insulator (SOI) transistors and copper wiring.
The addition of SOI alone can increase performance up to 20-30 percent beyond the use of copper by protecting the millions of transistors on a chip with a blanket of insulation, reducing harmful electrical leakage that wastes power. IBM was the first chipmaker to substitute copper for aluminum in chip wiring to boost performance and is now the first to perfect high-volume SOI manufacturing.
|
| HP strengthens research lab
|
|
|
Two Silicon Valley executives have joined the staff of HP Labs, the company's central research organization.
Kris Halvorsen, director of the Information Sciences Technology Lab at Xerox PARC, has been named director of the Solutions and Services Technologies Center. Patrick Scaglia, vice president of research at Cadence Design Systems, will be director of the Internet and Computing Platform Technologies Center, also reporting to Lampman. The center will focus on developing leading-edge hardware and software technologies for computing platforms, servers and storage systems.
|
| NASA to benchmark Star Bridge supercomputers
|
|
|
The U.S. National Aeronautics and Space Administration, Langley Research Center, and Star Bridge Systems have signed an agreement to jointly benchmark Star Bridge's technical high-performance supercomputers.
|
| Java 2 on Himalaya
|
|
|
Compaq Computer Corporation's NonStop Himalaya servers now provide full support for Java 2 Platform, Enterprise Edition (J2EE).
The addition of Java capability is a
step in Compaq's plan to advance the Himalaya platform as a core transaction engine for large-scale, Internet-driven e-business environments. Compaq also announced its new S74 family of faster NonStop Himalaya servers.
|
| 64-processor NUMA-Q E410 based on 700 MHz Xeon
|
|
|
IBM
introduced an
Intel-based server, the 64-processor NUMA-Q E410 and a two-way server, the Netfinity
3500 M20. The systems can run Windows and Linux. Powered by Intel's new 700 MHz Pentium III Xeon processors, the NUMA-Q E410 has shattered the industry's foremost data warehousing performance benchmark TPC-H, doubling
the previous results.
|
|
|
| US National Foundation for Functional Brain Imaging invests in Linux cluster
|
|
|
The US
National Foundation for Functional Brain Imaging (NFFBI) has invested in a 32-processor compute cluster based on the SGI 1400L server running
Linux. NFFBI, headquartered in Albuquerque, NM, is a research partnership with Harvard Medical School/Massachusetts General Hospital, University Of Minnesota School Of Medicine and New Mexico science and technical institutions such as the University of New Mexico, Albuquerque VA Medical Center, Sandia National Laboratories and Los Alamos National Laboratory. The system installed at NFFBI consists of eight SGI 1400L servers, each equipped with four Intel Pentium III Xeon processors.
|
| High-performance Linux cooperation by API, OCI and SCA
|
|
|
High-performance Linux provider Alpha Processor (API) announced separate collaborations with Object Computing, Inc. (OCI) and Scientific Computing Associates (SCA) that deliver a new level of bandwidth and reliability to hardware and software developers. This gives developers a set of
solutions that help them maximize performance on Alpha Linux.
|
| Linux on steady going
S/390 systems
|
|
|
With all the focus of high-end computing on highly parallel systems, we tend to forget much computing work is still done on fast S/390 based systems. IBM said it will make available Linux software and services for the S/390 enterprise server.
Linux for S/390 will be offered by SuSE and TurboLinux. IBM Global Services will offer complete Linux for S/390 services and support. The Linux for S/390 solution also includes IBM middleware, allowing portability and integration of applications and data on many different systems.
|
| Next gen
Intel workstations from SGI
|
|
|
SGI
introduced next-generation Intel processor-based workstations that deliver unparalleled system performance to creative and technical professionals running either Windows or Linux operating systems (OS). The systems ship with SGI VPro graphics also announced by the company today.
The new systems are based on Intel Pentium III and Pentium III Xeon processors and have been certified for the Microsoft Windows NT 4.0 and Red Hat Linux 6.1 operating systems.
|
| SGI releases compilers for Linux
|
|
|
SGI released
its SGI Pro64 compilers for the Intel Itanium processor to the Open Source Software Community. These Linux compilers contain additional optimizations that take advantage of the power of the Itanium processor over those of other public compiler implementations. This technology will support development of high performance IA-64 Linux applications.
|
| School on 'Linux for HPC'
|
|
|
The second annual MRCCS Summer School will be held, 4th - 15th September 2000, in Manchester, UK. It has as its theme the use of Linux for High Performance Computing. The Speakers are all international exports in their fields, including talks from SGI, Intel and the Portland Group. The practical sessions will include tutorials on performance, MPI and system administration as well as a hands-on 'Build your own Beowulf' competition.
|
|
|
| Space Imaging's satellite ground stations powered by Origion systems
|
|
|
Space Imaging has purchased six satellite ground stations powered by SGI Origin 2000
computer servers to capture and process the high-resolution imagery data transmitted to Earth by Space Imaging's IKONOS satellite, the first satellite to produce commercial high-resolution one-meter geospatial imagery of Earth.
The ground stations, developed by Raytheon Company, will be deployed at new Space Imaging regional affiliates overseas. Currently, SGI Origin 2000 servers are used in Space Imaging's primary ground station in Denver and at a Space Imaging regional affiliate ground station located in Greece.
|
| Visual Numerics to distribute software through
|
|
|
Visual Numerics
announced that
its PV-WAVE product is being distributed by San Bruno, Calif.-based
TestMart, an information and e-commerce market maker in the
test and measurement (T&M) industry.
All previously established sales channels for PV-WAVE,
including Visual Numerics' direct sales force and its network of international
distributors, remain intact.
|
| Veritas opens Toronto Visualisation Lab
|
|
|
Veritas Exploration Services opened its Canadian data visualization center. Resources include a Silicon Graphics Onyx 2 Infinite Reality supercomputer, high resolution digital rear-projection system, an 18' x 7' viewing screen, and an extensive tool kit of both third party and proprietary software applications. The facility is ideal for exploration teams to work within a collaborative setting offered by the main theater with seating for up to 30 geoscientists, and alternately in private visualization workrooms, allowing simultaneous analysis to be conducted by multi-discipline teams.
|
|
|
| Times N
introduces Processor Teaming technology
|
|
|
Times N Systems has developed a technology called Processor Teaming.
Processor Teaming i
offers an easy to use, powerful and scalable solution for compute and I/O intensive applications.
As companies grow and compute demands escalate, Processor Teaming gives a company control of its computing growth through the capability to efficiently scale most any increment of processors and I/O channels without having to upgrade to a Unix multi-processor server.
|
| New release of Cactus problem solving environment
|
|
|
The Cactus consortium has released a new version (beta 7) of its problem solving environment Cactus. Cactus is an open source problem solving environment designed for scientists and engineers. Its modular structure easily enables parallel computation across different architectures and collaborative code development between different groups.
Cactus has been tested on Linux, Cray T3E, Dec Alpha, HP Exemplar,
IBM SP2 and SGI Irix.
|
| Grid Tutorium in Mannheim
|
|
|
Following the Mannheim Supercomputer Seminar. a Tutorium on The Grid will be held at June 13 and 14.
The lectures include presentations on Globus,
benchmarking, clusters, resource mana gement, networking and diferent vendor solutions.
|
| Overview of recent supercomputers
|
|
|
What is a ccNuma machine? What is the clock cycle of the T3E? Can I still buy an SSP machine? How well do different supercomputers perform on standard benchmarks. As an expert in supercomputing and Grid computing, you know of course. However, I sometimes have difficulty to recall all of this useful detail information. That is why I look forward each year to the updated version of the "Overview of recent supercomputers" written by Aad van der Steen and published by the NCF. The 2000 version has just been published. Hence, I am up to date on all the issues concerning architectures of supercomputers. The report describes 16 current supercomputers from Avalon, Cambridge, Compaq, Fujitsu, Hitachi, HP, IBM, NEC, Quadrics, SGI, Sun and Cray Inc. Also included is a list of systems under development and systems that disappeared. Clusters are not included. However, the report notes the architectural trend in supercomputing seems to head in the direction of clusters of SMP nodes.
|
| LSF played crucial role in Athlon development
|
|
|
The development of the 1 GHz AMD Athlon processor was made possible through the use of Platform Computing's resource manager software LSF.
That according to members of the Austin, TX-based AMD team who made the world's fastest PC processor a reality.
|
| Cadre provides insight in petascale I/O patterns of
parallel systems
|
|
|
The University of Illionois inititated an US National Science Foundation funded project called Cadre. Cadre is a facility
to build a
base of empirical data on I/O access patterns of
high-performance computing systems. It will enable researchers and
educators to overcome limitations imposed by I/O on the overall performance of
petascale computing systems. Tools used
through the CADRE facility reflect
insight gained through development of the Pablo I/O characterization toolkit,
software tools for the performance analysis and optimization of parallel and
distributed systems, the umbrella project of which Csdere is
a part.
|
| AltaVista, Compaq and IBM Researchers create world's largest picture of the Web
|
|
|
Scientists from IBM Research, Compaq Corporate Research Laboratories and AltaVista Company have completed the first comprehensive "map" of the World Wide Web, and uncovered divisive boundaries between regions of the Internet that can make navigation difficult or, in some cases, impossible. The initial findings were
be presented lat week simultaneously at the 9th International World Wide Web Conference, Amsterdam
and at the ACM PODS 2000 Conference, Dallas.
|
| The Information Power Grid is changing our world
|
|
|
The new computational grid, short "the Grid" is a getting a lot of attention. Combining a lot of techniques
to
make a world wide seamless accesible computing resource out off all teh systems available over the Internet is the ultimate goal. What is the current state-of-the-art? Which hurdles are there to take. Wolfgang Gentszch, who, after a merger,
changed the name of his company to Gridware, presented the key-note at HPCN Europe 2000 on this subject. Apart from new technologies, it will also be the abundance of networking bandwidth availble shortly, that will allow access to computing for everyone,
anywhere, anytime.
|
| Futures on computing
|
|
|
Futures in trading stocks on for instance the stock market are well know: you sell something
that you do not yet have, or buy something to be delivered in the future. This concept has know also be introduced in computer services by Australian researchers H. James and K. Hawick at the University of Adelaide and presented at the HPCN Europe 2000 in Amsterdam. It makes their computing services that are run all over the country run smoother and more effecient. Their Gid system, called DISCworld, is especially designed to hide latency
that will inevitably pop up in computing systems that are distributed in a wide area network over cities or regions. Their DISCworld remote access mechanism futures, provide a mechanism to reference data that may not be available, but that will be in the future. This reference can than be used by an application, and even passed on for further processing.
|
| Hundred e-mails to get one high-speed cross-atlantic network connection for a metacomputer
|
|
|
At SC99 a metacomputing experiment
was aimed a connecting supercomputers in Japan, the USA, Germany and the UK, to provide one single big system that could be made available to one single application that would then use all this power to solve problems never solvable before. This Grid
experiment avant la lettre
is on the bleeding edge of technology. Hence lessons can be learned for the future of the Grid. Fumie Costen, of the Manchester Computing Centre explained the experiment at HPCN Europe in Amsterdam.
|
| Version
2.0.2 of Scali's SSP software
|
|
|
Scali released version 2.0.2
of its SSP -
Scali Software Platform - fixing some problems. SSP is
avialable on Linux (Redhat and Suse) and Solaris.
|
| 64 processor SGI 2800 at Mississippi Center for Supercomputing Research
|
|
|
The Mississippi Center for Supercomputing Research (MCSR) at the University of Mississippi has installed
a 64-processor SGI 2800 server. The center's largest single application is the use of software from Gaussian, Inc. of Carnegie, Penn.
for molecular modeling studies. MCSR upgraded its eight-processor SGI Origin 2000 system to
a 64-processor configuration with 16 GByte of memory, primarily to make Mississippi researchers more competitive nationally in the USA in the solicitation of contracts and grants.
|
| International Conference on Cluster Computing
|
|
|
The second call for the International Conference on Cluster Computing
in has been issued. The conference will take place November 28 - December 2, 2000 in Chemnitz, Germany.
|
| NVIDIA selects
Gridware's Codine/GRD
|
|
|
The
Codine/GRD
computing resource management software has been selected by NVIDIA
as the foundation for the tightly-clustered, heterogeneous computing environment it is building to handle the EDA processing requirements for its next generation of 3D graphics processors.
|
| Intel introduces new processors for high-end servers
|
|
|
Intel Corporation
began initial shipments of its next generation Pentium III Xeon processors that deliver new levels of performance to high-end, Intel-based server platforms based on 4- and 8-way multiprocessing.
The new 700 MHz processors are the first "large cache" Pentium III Xeon processors based on Intel's advanced 0.18-micron manufacturing technology, offering 1 or 2 Megabytes
of "on-die" level-two (L2) cache. The 2 MByte version integrates a record 140 million transistors on a single microprocessor die.
The clock speed increase, along with an enhanced cache design, boosts overall system performance by 13-46 percent, depending on applications and configurations.
|
| Aachen release new version of MP-MPICH
|
|
|
The University of Aachen, Germany, has released version 1.2.0 of
MP-MPICH (Multi-Platform MPICH). It includes a number of new features for SCI-MPICH and NT-MPICH as well, like MPE logging API, MPI-IO for NTFS and reliable remote execution.
|
| Cornell Theory Center acquires 128 processor Dell cluster
|
|
|
The Cornell Theory Center (CTC)
acquired
a new 128-processor, high performance cluster. The new Velocity+ cluster, funded through CTC's Advanced Cluster Computing Consortium (AC3), is the latest Windows 2000-based Dell/Intel/Microsoft cluster to be installed at CTC. Velocity+ will be dedicated to strategic, computationally challenging applications, aimed at solving highly complex problems in fields such as molecular modelling and multiscale materials simulation. One of the strategic applications lined up to run on Velocity+ comes from CTC's Parallel Processing Resource for Biomedical Scientists, which is funded by the National Center for Research Resources.
|
|
|
| Combining information on supercomputers from several sources with XML - an example using the TOP500 and the "Overview of recent Supercomputers"
|
|
|
There are many sources of information on supercomputers available. One of the most important ones is the TOP500 list of most powerful supercomputers in the world. It contains data on the position in the TOP500, the performance and where the machine is located. Another source is the "Overview of recent supercomputers" that contains a description of the supercomputer systems available, with system parameters, and benchmark information. With XML-technology we have combined information from the two sources to generate a new document, containing a description of each architecture, extended with a sub list of the TOP500 on that machine. The new document is generated with XSL-Transformation-stylesheets, that leave the original sources untouched. When there is a new version of for instance the TOP500, running the stylesheet is enough to generate a new, updated version of the assembled information document.
|
