The high performance computing (HPC) system also will play a vital role in supporting additional departments at NTNU by providing a substantial increase in calculation capability so that these Centers of Excellence may stay competitive on an international level.
"NTNU and met.no have worked together for more than 20 years on high performance computing systems for research and numerical weather prediction", stated Roar Skalin, director of Information Technology at the Norwegian Meteorological Institute. "With this newest SGI Altix ICE solution, the compute resources available for our numerical weather prediction will increase by a factor of 20."
The system will consist of 13 next-generation Altix ICE cabinets performing at over 275 teraflops, along with four SGI InfiniteStorage 16000 racks providing 1.2 petabytes of storage using the Lustre file system. This configuration will allow met.no to use fine scale operational atmospheric and oceanographic models to dramatically improve weather forecasting for Norway and the adjacent seas. The new system will allow the institute to run operational atmospheric models at a resolution of 2.5km or better, and in parallel allow the researchers to increase the usage of satellite observations.
"With this SGI technology collaboration, the Norwegian University of Science and Technology and the Norwegian Meteorological Institute will be able to greatly improve their weather forecasting capabilities, as well as vastly accelerate time to insight for more standard applications common to HPC academia", stated Rod Evans, vice president of SGI EMEA. "The flexibility and scalability of this next generation solution also creates the foundation for an HPC system that can continue to grow without disruption to their service as their requirements change and expand."
The Altix ICE platform has been purposely built for high-performance technical computing. The latest generation being purchased by NTNU adds even greater flexibility, with a broader range of compute nodes, topology types, InfiniBand fabric design, power , density, and cooling. Altix ICE can scale seamlessly from 10 teraflops to 100 petaflops without service interruption due to its unique architecture created with the hierarchical management design, topology options, and integrated InfiniBand switches that replace large, inflexible core switches.
"The ability to quickly and accurately forecast the weather is critical to citizens and businesses of Norway and surrounding countries", stated Rajeeb Hazra, general manager of High Performance Computing at Intel. "We are pleased to work with SGI to provide an HPC solution powered by the next generation Intel Xeon processor series that can intelligently adjust performance to data-intensive applications like those used by the Norwegian University of Science and Technology (NTNU) and the Norwegian Meteorological Institute."