October 13, 2006
CSC Finland, the Finnish IT center for science, will acquire a Cray massively parallel processing (MPP) system delivering over 70 teraflops of compute power to CSC's high performance computing users. The new Cray supercomputer, code-named Hood, will be installed in stages beginning later this year and continuing through 2008. It will replace a four-year-old cluster system that can no longer keep pace with the performance needs of the Finnish research community, which is currently doubling its computer usage every 16 months. CSC will also be upgrading its cluster infrastructure with an HP ProLiant system. The purchases are being funded by Finland's Ministry of Education.
"We selected the Cray Hood supercomputer after an extensive acquisition process that involved surveying 35 different research groups, closely analyzing the available technologies and benchmarking competing systems," said Kimmo Koski, managing director of CSC Finland. "Our goal was to procure the most powerful system for the funds that we had available. The new Cray supercomputer will provide the capability required by our diverse research groups and bring Finland back to the leading edge in Europe."
CSC provides IT infrastructure, skills and specialist services for universities, polytechnic colleges, research institutions and companies across Finland, and collaborates with various research institutions worldwide. The new Hood system will be used for research in areas such as physics, chemistry, nanotechnology, linguistics, bioscience, applied mathematics and engineering. The acquisition project was started in 2005 based on the budget proposal of Finland's Ministry of Education and the Council of State. During the acquisition process, the offered systems were tested by running the CSC benchmark set, which consists of codes used by CSC's customers.
"We are delighted to welcome CSC Finland as a Cray customer," said Ulla Thiel, vice president of Cray Europe. "As one of the most prominent supercomputing centers in Europe with a broad range of high performance computing disciplines, CSC is in an excellent position to maximize the advantages of our new Hood MPP system in solving the most demanding scientific problems and exceeding the requirements of a production supercomputer environment."
Hood is the successor to the Cray XT3 supercomputer, which has set new performance records and was named Product of the Year by Scientific Computing & Instrumentation in 2005. The Hood architecture is designed for high levels of scalability and sustained performance on a wide range of scientific applications. It employs thousands of AMD Opteron processors running a lightweight operating system tuned to optimize sustained application performance. Processors are interconnected on a high-bandwidth, low-latency communications network custom designed for scale by Cray, based on HyperTransport technology and the Cray SeaStar routing and communications chip.
In addition to the new Cray acquistion, CSC will be extending its super-cluster system as well. The new super-cluster system is an HP CP4000BL ProLiant cluster running the HP XC cluster software stack, integrated with a 100 TB HP Scalable File Share high-performance storage system based on Lustre technology. This system uses the recently introduced cClass blades from HP, incorporating a 4X DDR Infiniband interconnect. The cluster contains 2048 2.6 GHz AMD Opteron compute cores, 4 TB of memory, and has a peak performance of 10.6 Tflops.
The super-cluster system increases the throughput capacity for users who need moderate numbers of processors but large memory and high-performance storage. Thanks to the high-speed interconnect, it can be used for small or medium-size parallel jobs as well. The system will be one of the most powerful installations of HP in Europe.
"HP looks forward to a productive collaboration with CSC to help accelerate the productivity of scientific research in Finland," said Martin Walker, HP Segment Manager for Scientific Research, "The industry standard cClass infrastructure accommodates next generation processors, memory, and interconnect, making it possible for CSC to remain on the leading edge of computing technology."
"AMD is proud of being able to contribute to such an exciting project that will boost the scientific research in Finland for a long time," said Francesco Torricelli, worldwide manager of high performance computing at AMD. "The AMD Opteron processor provides the scalability, performance, I/O bandwidth and power efficiency IT managers in large scientific research centers are looking for. Furthermore, the AMD64 architecture inherent in the AMD Opteron processor allows an easy and seamless upgrade path from dual- to quad-core processors. All within the same power specifications and infrastructure -- enabling the researchers to easily scale with their computing needs."
The allowances available for the CSC upgrades are 10 million Euros and the vast majority of it will be used for the supercomputer and super-cluster. A minor portion will be used for a new data storage system. Splitting the order between two various systems makes it possible to differentiate CSC's computing services according to the customer needs.
The Cray supercomputer will be dedicated to the most challenging problems in HPC and allow applications to scale to several thousands of processors while the HP super-cluster offers good price/performance ratio for jobs of moderate parallelism needing large memory or plenty of disk space.
The new resources will have a major impact on the computational research in Finland. The nanoscientists, who are the biggest users of CSC's resources in terms of CPU-time, as well as other large groups, including environment researchers, chemists, bio-scientists and physicists will all certainly be able to benefit from the large increase in computing power. Half of the centers of excellence in research, nominated by the Academy of Finland, are CSC's customers and use one third of the computing capacity.
"One of the most rapidly growing areas of research and product development today is nanoscience and -technology, which utilizes atom-level scientific understanding to build up new kinds of functional materials and devices. Nanoscience thus relies on understanding complicated atomic interactions, and the best way to obtain that is using massive supercomputing capability." says professor Kai Nordlund from the University of Helsinki. He continues: "The new capacity will enable, for instance, studying dynamic processes in entire nanoobjects on the quantum level, something which very few research groups yet can do anywhere in the world."
"Climate system models supply Finnish society with information on climate change. These models describe the atmosphere, oceans and biosphere with all their mutual interactions, making them computationally and expert-wise very demanding. Need for computational resources increases in pace with the higher resolution, which is necessary for modeling of local and short-term weather extremes," says research professor Heikki Järvinen from the Finnish Meteorological Institute (FMI).
Professor Järvinen emphasizes that the new supercomputer capacity at CSC facilitates the climate research at FMI and in the universities to support preparation of national climate policy and to evaluate the human impact on climate on a national scale.
Source: CSC Finland and Cray Inc.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?