Researchers and scientists at DWD will utilize the petascale systems to produce higher resolution and more accurate global and regional weather predictions to help fulfil the organisation's wide array of weather responsibilities, including the meteorological safeguarding of aviation and shipping and the issuance of official warnings of weather occurrences that could become dangerous to public safety.
"At DWD it is our mission to ensure the highest levels of accuracy in weather and climate information services", stated Dr. Dieter Schröder, CIO at DWD. "Our researchers and scientists must be able to rely on their supercomputing systems for high availability, performance and scalability, so we needed a workload manager proven to operate in large, complex production environments. PBS Professional provides the bulletproof reliability and flexibility we need to manage our critical numerical weather prediction workloads and ensure faster, more reliable computing performance."
"Everyone relies on weather prediction, from first responders to vacationing golfers", stated Bill Nitzberg, chief technology officer of PBS Works at Altair. "Delivering high-fidelity, accurate and timely forecasts requires a level of reliability and performance far above what is required in other HPC areas. Reliability is a key strength for PBS Professional, which is why top global weather centres like DWD choose Altair for the workload management capabilities required to deliver their world-class services."
A longtime leader in HPC workload management, Altair's PBS Professional has been used for more than 20 years at thousands of customer sites to easily share distributed computing resources across geographic boundaries. PBS Professional is the flagship product for PBS Works, a comprehensive suite for optimizing HPC and Cloud computing environments including portal products for job management, analytics and remote visualization.
Numerical weather prediction uses many different types of simulations, modeling techniques and components at different spatial and temporal scales. The meteorological models used are extremely data-intensive, and the computational and data requirements to simulate these models are massive. In addition, with increases in available data and better understanding of the requirements for improving forecast accuracy, demands for handling complexity and big data movement grow even higher. Thus, weather centres require not only scalable architectures that support highly compute-intensive applications and efficient data movement - they also require a well-integrated solution built by experts who understand the experience of architecting and delivering systems at scale and complexity.
Leaders in the field of HPC solutions for weather and climate centres, Cray and Altair have collaborated to develop an integrated solution offering a breakthrough in system utilization strategy and high-performance workload scheduling and management. Designed with the demands of numerical weather prediction in mind, the Cray-Altair solution enables weather facilities to exploit peak performance capabilities while improving service quality and utilization rates. The solution is described in a white paper published by Altair and Cray and available at http://www.pbsworks.com/cray-weather-wp .
"Cray and Altair have a longstanding partnership, and we are pleased that the combination of the Cray XC30 supercomputer and PBS Professional has proven to be a compelling solution for DWD's meteorological workloads", stated Peg Williams, senior vice president of high performance computing systems at Cray. "PBS Professional is a proven workload management product and an excellent choice for organisations and businesses with large, complex systems where uptime and reliability are top priority."
You can read more about Cray and Altair's solutions for numerical weather prediction at http://www.altair.com/cray-and-altair . You can learn more about PBS Professional at http://www.pbsworks.com/Product.aspx?id=1 .