SCIENCE
Adaptive Computing Selected to Manage Job Scheduling for Sweden’s Leading Academic HPC System
Adaptive Computing has announced that Sweden’s Royal Institute for Technology (KTH) has chosen Moab Cluster Suite to manage the organization’s Center for High Performance Computing (PDC). With Adaptive Computing’s flexible technology, PDC can now manage the additional complexity of an expanded HPC network while supporting future initiatives such as integrating and managing disparate server and OS architectures.
“The upgrade of our existing Cray XT6m system to a new Cray XE6 system, with three times the capability of the current system, means we require a workload management solution that will keep up with the growing complexity of the PDC high-performance computing and storage services,” said Dr. Erwin Laure, director of PDC-HPC at KTH. “Moab is an industry-proven technology used in some of the world’s largest HPC systems. As an intelligent metascheduler, Moab will schedule jobs for optimal efficiency, based on defined policies, service-level agreements, and current and projected workload. These advanced capabilities made Moab a clear leader in our selection process.”
PDC operates leading-edge, high-performance computers and offers easily accessible computational resources that cater to the needs of Swedish academic research and education. PDC is involved in major international projects to develop high-performance computing (HPC) for the next ten years and beyond. The primary PDC system is a 305 teraflops Cray XE6 that will integrate with the Partnership for Advanced Computing in Europe (PRACE) network. PRACE is a pan-European research infrastructure for high-performance computing, allowing researchers from across Europe to apply for time on high-performance computers via a central peer-review process. Additional HPC and HTC systems at PDC add to an integrated total capacity of almost 500 teraflops.
“Moab Cluster Suite provides PDC with the flexibility required to satisfy Sweden’s diverse HPC user community while optimizing the throughput of its HPC infrastructure,” said Michael Jackson, president of Adaptive Computing. “Adaptive Computing is proud to support the management-level requirements of one of the most powerful supercomputers in Europe.”