Hewlett Packard Enterprise (HPE) is building a new supercomputer for the United Weather Centres – West (UWC-West), a collaboration between the Danish Meteorological Institute, Icelandic Met Office, Met Éireann, Ireland’s national weather service, and the Royal Netherlands Meteorological Institute to advance weather forecasting for the four nations.
The supercomputer, which will be based in an Icelandic Met Office data center facility and powered by local renewable energy, will be built using the HPE Cray system that features powerful, end-to-end performance to speed time-to-predictions with higher resolution, helping weather services issue timely alerts and improve public services.
By combining national resources and learnings on the shared supercomputer, the four meteorological services will improve weather modeling to generate more detailed forecast updates and make predictions every hour. The overall advanced performance will also enable research into extreme weather events, such as heatwaves, heavy rain, flooding, and snowstorms, which are more often triggered by climate change. This will help provide timely warnings to alert local governments and communities to prepare for severe conditions and protect human lives, livestock, and property.
“As European nations continue to face challenges with new, dynamic weather patterns caused by climate change, weather forecasters will need powerful high performance computing (HPC) capabilities to evolve weather models and simulate vast amounts of complex data to unlock accurate, real-time forecasts,” said Bill Mannel, vice president and general manager, HPC, at HPE. “We are pleased to see Denmark, Iceland, Ireland and The Netherlands take action and join forces in the United Weather Centres- West (UWC –West) to strengthen weather services in Europe. We are honored to have been selected to design their new supercomputer, using end-to-end HPC technologies, and support their multi-national mission for advancing weather forecasting.”
New UWC-West Supercomputer Fuels Significant Leap in Weather Forecasting
UWC-West’s new supercomputer will be comprised of two systems, one dedicated to operational weather forecasting and another for broader weather and climate research, helping to harness complex atmospheric and oceanic data that helps model and simulate weather at a higher resolution.
“Our countries have a long history of working together in weather forecasting – often the weather experienced in Ireland or Iceland today is the same weather experienced in Denmark and The Netherlands tomorrow,” said Marianne Thyrring, the UWC-West Chair “The UWC-West supercomputer is the first step in a powerful collaboration between weather services in Europe, and it is vital that we continue working closer together to improve our weather forecasts and understanding of how climate change will impact our countries.”
The new system will specifically support a next-generation Numerical Weather Prediction (NWP) model, called HARMONIE-AROME, which has been developed under the HIRLAM-ACCORD collaboration, a consortium now comprised of 26 national meteorological services across Europe and North Africa that is aimed at improving short-range weather predictions.
Built using the HPE Cray system, UWC-West’s new system will also deliver the following powerful, end-to-end HPC technologies to support processing, storing and management of vast amounts of data:
Addressing demands for higher speed and congestion control for larger data-intensive and AI workloads with HPE Slingshot, the world’s only high performance Ethernet fabric designed for HPC and AI solutions
Expanded storage to support and share complex workloads in modeling, simulation and AI using the Cray ClusterStor E1000 storage system from HPE and HPE Data Management Framework
Enabling fine-grained centralized monitoring and management for optimal performance with the HPE Performance Cluster Management, a system management software solution, and a fully integrated software suite to optimize HPC and AI applications using the HPE Cray Programming Environment
