 Company Our Offices Netherlands Office
ClusterVision BV
Nieuw-Zeelandweg 15B
1045 AL Amsterdam
Netherlands
Tel: +31 20 407 7550
Fax: +31 84 759 8389
UK Office
ClusterVision Ltd
12 Westgate House
The Island
Gloucester GL1 2RU
United Kingdom
Tel: +44 870 080 1990
Fax: +44 870 132 4332
Germany Office
ClusterVision Deutschland
Feringastraße 6
85774 München
Germany
Tel: +49 899 921 6433
Fax: +49 899 218 5311
France Office
ClusterVision France
Tour Ariane
5 Place de la Pyramide
92088 Paris La Défense
France
Tel: +33 1 556 81080
Fax: +33 1 7772 5634
Italy Office
ClusterVision Italy
Via Senigallia 18/2 - Torre A
Milano 20161
Italy
Tel: +39 02 6467 2654
Fax: +39 02 4438 6343
Switzerland Office
ClusterVision Switzerland
18, Avenue Louis-Casaï
1209 Geneva
Switzerland
Tel: +41 22 747 7825
Fax: +41 22 594 8468
Nordics Office
ClusterVision Scandinavia
Nydalsveien 33, Torgbygget
0484 Oslo
Norway
Tel: +47 21 52 0089
Fax: +47 21 54 7448
Spain Office
ClusterVision Spain
Pinar, 5
Madrid 28002
Spain
Tel: +34 91 745 6874
Fax: +34 91 791 5338
| 
 MAX PLANCK INSTITUTE GÖTTINGEN
Simulating Proteins at the Max Planck Institute for Biophysical Chemistry
The Max Planck Institute for Biophysical Chemistry in Göttingen is renowned in the field of theoretical and computational biophysics and uses several clusters supplied by ClusterVision for their research into the dynamics of proteins and other biological macromolecules.
The Max Planck Institute in Göttingen has purchased several clusters from ClusterVision in order to acquire the compute power required for their protein simulations. The research on these powerful supercomputers has been essential for many of the scientific publications of the theoretical and computational biophysics department.
The clusters are all based on dual processor servers with processors from Intel® and AMD. The clusters run the ClusterVisionOS™, and multiple master nodes are used to monitor and manage the clusters and distribute the slave node images.
|
"Since our first purchase in 2004, ClusterVision has repeatedly impressed us not only with their high standard in quality and reliability, in both hard- and software, but also with their flexible and competent handling of reclamations, in the few cases of technical problems we've faced thus far." — Dr Bert de Groot
|
| INRIA ROCQUENCOURT
ClusterVisionOS™ Installed on INRIA Clusters
The Institut National de Recherche en Informatique et en Automatique (INRIA) and Dell France asked ClusterVision to install its Linux-based cluster management, monitoring and administration environment, the ClusterVisionOS™, on a series of new and existing clusters.
INRIA is France's national research institute for information and communication sciences and technologies. In order to help research projects which occasionally need access to large computing power, or in complement to their own computing power, the SED (Service d'Expérimentation et de Développement) of INRIA Rocquencourt has put in place a mutualised calculation platform. This platform is composed today of multiple clusters, which can be used by research projects for all types of intensive calculations, be they parallel or not.
Working with Dell France, ClusterVision has installed an AMD Opteron™ based cluster at the INRIA in Rocquencourt. This cluster consists of a Dell PowerEdge™ 6950 master node and 32 Dell SC1435 slave nodes, and will reinforce the existing cluster infrastructure of the SED. The clusters currently have about 80 users, spread over 20 projects. The support that the SED offers is on one part the possibility of reconfiguration of systems and networks, and exclusive reservation of "sub-clusters". On
the other part, the installation of libraries and compilers, and linking these with activities.
|
"We were very impressed with ClusterVision's cluster management software stack: ClusterVisionOS™, and decided to ask ClusterVision to install it on all our other existing clusters. We have a relatively complex user base. The modularity offered by the ClusterVisionOS™ thus met and exceeded our needs."
— Mr Pierre-Louis Neumann.
|
| UNIVERSITY OF CAMBRIDGE
Europe's Fastest x86 Cluster at the University of Cambridge
Built by ClusterVision Using Dell Intel Servers and InfiniPath InfiniBand Interconnect
ClusterVision and Dell have completed the installation of Europe's
fastest x86 cluster for the High Performance Computing Facility
at the University of Cambridge. Built using compute servers and storage
units from Dell and processors from Intel®, the supercomputer
will provide a central compute and storage facility to all scientists of
the University of Cambridge.
More than 1152 dual-core Intel® Xeon® processors
provide the University of Cambridge
with more than 27 TeraFLOPs (27 x 10^12
calculations per second) of raw computing
power. This entered the cluster at position 20 of the
TOP500 list of fastest
supercomputers in the world.
|
"ClusterVision's role has been key in rapidly turning the Dell-supplied hardware into a usable and manageable cluster, ready for the TOP500 benchmarks." — Dr Paul Calleja, Director of HPC, University of Cambridge
"ClusterVision played a key part in the installation and integration of the supercomputer, and will be providing first-line support services to Cambridge University." — Mr Iain Campbell, General Manager Public Sector, Dell UK
|
| NUMECA INTERNATIONAL
NUMECA Uses ClusterVision's Compute Clusters for Computational Fluid Dynamics
Several compute clusters based
on Intel® Xeon® architecture help
NUMECA International to develop
and test their broad range of Computational
Fluid Dynamics (CFD)
software and to run extensive
simulations for their customers.
Air flow around aircraft, cars and buildings;
fluid and gas flow through pumps,
compressors, turbines and around
yachts are but a few of the applications of
Computational Fluid Dynamics (CFD). NUMECA
International is a leading provider of
CFD software and consulting services worldwide
and has chosen ClusterVision to supply
its in-house compute clusters to develop their
range of CFD programs and to run simulations
for customers that do not have the
compute power, skills or software to perform
those simulations.
NUMECA's clusters are based on 32-bit
and 64-bit Intel® Xeon® processors, combined
with Gigabit Ethernet and InfiniBand networks
for high-bandwidth and low-latency
communication. The clusters together provide
more than 1.5 TFLOPs of raw compute
power. The cluster runs the Linux-based
ClusterVisionOS™ cluster operating system
and software environment which includes all
software required to effectively use and manage
the clusters.
|
"Our partnership with ClusterVision opens the door to large-scale CFD simulations within short industrial time frame. The new cluster will bring large additional calculation power to NUMECA and support our custome and partner demands to run challenging multi-physics and multi-disciplinary simulations."
— Mr Marc Tombroff
|
| NEDSTAT
ClusterVisionOS™ Manages Nedstat Server Farm
European market Leader in Website Analytics Chooses ClusterVisionOS™
One of the Netherlands' most successful
dotcom companies and European
market leader in website
analytics, Nedstat (nedstat.com),
has chosen the ClusterVisionOS™
cluster operating system and
software environment to run and
manage their new server farm in
Amsterdam.
As websites become more and more
central to businesses' communication
and online business, the importance
of reliable and comprehensive website usage
statistics is evident. This is where Nedstat
excels and has gained a leading position in
Europe as provider of website analytics. Its
Sitestat software provides detailed statistics
and analysis of when and how visitors access
websites and thus enables Nedstat customers
to improve the effectiveness and profitability
of their online communication and business.
Nedstat's server farm was growing rapidly so there was a need for sophisticated software
to manage and monitor the farm and take care of the distribution
of different software images to the various servers.
The ClusterVisionOS™ provided the required
functionality and allowed Nedstat developers
to focus on Sitestat while ClusterVision
engineers took care of the operating system
and management software on the server
farm.
|
"We considered different options for the management software on our new server farm, but were very impressed with the ClusterVisionOS™. It has all the functionality we require to effectively run our server farm and monitor its status. Its flexible image management capability is particularly useful for us as it allows us to rapidly deploy new versions of Sitestat to different sections of the server farm." — Mr Michiel Berger
|
| UNIVERSITY MEDICAL CENTRE UTRECHT
Analysing the Brain Using a ClusterVision Compute Cluster
A compute cluster supplied by ClusterVision helps scientists at the University
of Utrecht (NL) to analyse the human brain and better understand
psychiatric diseases.
The cluster based on Intel® Xeon® processors
supplied by ClusterVision helps to analyse
thousands of MR images to find genetic
and environmental markers for variance in human
brain structure in health and psychiatric
disease.
The analysis consists of dividing the
image into meaningful
parts, the so-called
segments. In the segmentation
process,
the voxels ("volume
elements", similar to
pixels of a digital photo)
of the image are
classified as belonging
to a certain tissue
type or structure. The
segments can be large
structures such as the
whole brain, or structures as small as, for instance,
the amygdala and the third ventricle.
Volumes are obtained by counting the number
of voxels in the segment and multiplying this
by the volume of one voxel.
In order to obtain
the desired statistical power, often large
numbers of subjects have to be scanned and
the MR scans segmented. Automation of the
segmentation steps leads to objective analysis
and less time consuming processing.
|
"The cluster supplied by ClusterVision has really helped us to speed up the processing, analysing and storing of the thousands of MR images we use for our research. We are very happy with the way the cluster works. The cluster provides a reliable resource for all scientists in our research group" — Dr Hilleke Hulshoff Pol, Head of the Structural Neuroimaging Section
|
| ALBERT EINSTEIN INSTITUTE
Investigating Einstein and Beyond &mdash Albert Einstein Institute in Potsdam Receives Opteron/InfiniBand Cluster
The Max Planck Institute for Gravitational Physics (Albert Einstein Institute)
has received a new ClusterVision cluster for research in Einstein's
General Theory of Relativity. The HPC cluster, named Belladonna, will
be used in the simulation of gravitational-wave sources coming from
compact astrophysical objects such as black holes, neutron stars and
supernova explosions.
The cluster supplied by ClusterVision includes
more than a hundred dual-core
AMD Opteron™ processors which are
interconnected using a high-performance
InfiniBand network. It will
provide the computational
power needed to solve to
high accuracy Einstein's
equations and study
processes such as the
dynamics of a system
of two black holes orbiting
around each other
and gradually inspiralling
as a consequence of the emission
of gravitational waves.
|
"The prospects in numerical relativity have never been as exciting as they are now. The solutions computed with Belladonna will help lay a bridge between numerical relativity and gravitational-wave data analysis"
— Professor Luciano Rezzolla, Head of the Numerical Relativity Group at the Albert Einstein Institute
|
| DAS-3 GRID
ClusterVision Builds DAS-3 Grid for Dutch Universities
ClusterVision has been selected to build the DAS-3 grid (Distributed
ASCI Supercomputer) in the Netherlands. DAS-3 will consist of five
Linux supercomputer clusters, with an aggregate theoretical peak
performance of more than 3.8 Teraflops.
The individual clusters will be hosted at four leading Dutch universities and will be connected with SURFnet's dedicated multi-colour optical network and Myricom's
Myri-10G interconnect. Because of the advanced
interconnect technology, data transfer
rates between clusters will be up to 80
Gigabits per second (Gb/s).
DAS-3 is third-generation DAS. Unlike its
DAS-1 (1997) and DAS-2 (2002) predecessors,
DAS-3 will use the inter-city SURFnet optical
network as the grid's backbone. In total, DAS-
3 will link more than 550 AMD Opteron™
processors, 1TB of memory and 116TB of mass
storage. The five grid clusters will be installed
with the Linux based ClusterVisionOS™ cluster
operating system and software environment.
All the DAS grids were designed as research
environments used for studying distributed
computing architectures. Topics of
interest include parallel programming languages,
operating systems research, runtime
language systems and algorithmic research.
The Ibis open source Java grid software environment
has been studied extensively using
the DAS-2 grid.
Compared to other grids, the DAS architecture
is very homogeneous in nature. Although
each individual cluster's memory capacity
and number of processors may vary, all
systems are Opteron based, running the same
Linux OS and linked with the same interconnect
hardware. This
greatly simplifies system
administration.
More importantly,
since the fundamental
characteristics of each
system are identical,
distributed application performance is much
simpler to measure. No "apples to oranges"
comparisons are necessary.
|
"When this thing goes operational in August it's going to be the fastest grid of clusters in the world" — Dr Chuck Seitz, Myricom CEO
|
| UNIVERSITY OF DORTMUND
LIDO GEHT LOS! &mdash Teraflops Linux Cluster for the University of Dortmund
The University of Dortmund has recently inaugurated its new central
supercomputer: Linux Cluster Dortmund (LiDO). ClusterVision
delivered the cluster earlier this year to the High Performance
Computing centre of the university.
LiDO will be used in
leading-edge scientific
research by scientists
from the University
of Dortmund, Fachhochschule
Dortmund and other
universities in the state
of Nordrhein Westfalen.
With more than 420 AMD
Opteron™ processors, 1.3TB of memory, 26TB
of disk storage and a peak performance of
more than 1 TFLOP, it is 55 times more powerful
than its predecessor, an IBM SP computer
installed in 1996.
"It was very impressive for us to see how ClusterVision passed successfully all scientific benchmark configurations which had been designed by scientists in Dortmund to directly test the potential of the new
computer system for highly sophisticated
technical simulation methods,
for instance for Computational Fluid Dynamics
(CFD). So, we are very sure that we
have found our new powerful work horse
for solving grand challenge scientific problems
in the future in our Virtual Numerics
Labs", said Prof. Stefan Turek, Vice-Dean of
the Department for Mathematics and Head
of the Institute for Applied Mathematics, University
of Dortmund.
|
"We evaluated a large number of proposals from many leading IT vendors and decided on ClusterVision because of their excellent reputation and the outstanding price / performance ratio. ClusterVision also best understood and fulfilled our strict requirements" — Jörg Gehrke, Division Leader Server and HPC, HRZ, University of Dortmund.
|
| CERN
850 CPUs at CERN — ClusterVision Delivers More than 425 Dual Intel® Xeon® Servers to the World's Largest Particle Physics Laboratory
CERN has taken delivery of more
than 425 dual Intel® Xeon® servers
from ClusterVision earlier
this year. ClusterVision won the
orders in competitive tender
amongst many competitors from
across Europe.
The servers will be used to analyse data produced by the Large Hadron Collider (LHC) — the largest scientific instrument
currently being built on this planet. When the
LHC begins operations in 2007, it will produce
roughly 15 Petabytes (15 million Gigabytes) of
data annually, which thousands of scientists
around the world will access and analyse. The
LCG (LHC Computing Grid) project has as
mission to build and maintain a data storage
and analysis infrastructure for the entire high
energy physics community that will use the
LHC.
ClusterVision is no stranger to the LCG
project, having already supplied to many
of the participating research institutes and
universities, including the UK National Grid
Service (NGS), Rutherford Appleton Laboratory,
NIKHEF, the University of Lancaster (see
picture) and Queen Mary College.
|
| | |
|
