Energy-efficient supercomputer on the way for NREL

Technology company Hewlett Packard Enterprise (HPE) has announced it is building a new supercomputer for the National Renewable Energy Laboratory (NREL), said to be more energy efficient and 3.5 times more powerful than its existing system.

The new development is part of a longstanding collaboration between HPE and the US Department of Energy (DOE) to apply advanced supercomputing and HPC solutions to accelerate research across various DOE agencies. The new, fast-performing system, which NREL has named Eagle, will run more detailed models that simulate complex processes, systems and phenomena to advance early research and development on energy technologies across fields including vehicle, wind power and data sciences.

Claimed to be the world’s largest HPC system dedicated to advancing renewable energy and energy efficiency technologies, Eagle is powered by the HPE SGI 8600, a system designed from the ground up to run complex HPC workloads at petaflop speeds — or a quadrillion (thousand trillion) floating point operations per second (FLOPS). Additionally, with the HPE SGI 8600, Eagle is gaining a warm liquid cooling system that captures 97% of wasted heat to re-use in other areas of its hosted facility like surrounding office space and labs.

“We are strongly committed to architecting technologies that power the next wave of supercomputing and are creating advanced HPC systems, while scaling energy efficiency in data centres, to get us there,” said Bill Mannel, Vice President and General Manager, HPC and AI Group, HPE. “Through Eagle and our overall ongoing collaboration with the US DOE and NREL, we are advancing research to bolster innovation in energy and sustainability.”

HPE is enabling Eagle with a fully integrated, turnkey system that includes advanced next-generation compute, network and storage capabilities. The system runs on the Intel Xeon Scalable processors, uses Mellanox EDR InfiniBand fabric and comprises a total of 76,104 compute cores and 2144 dual-socket computes nodes, each with memory ranging from 96 to 768 GB, delivering a peak performance of 8 petaflops.

“With Eagle, we are gaining significant compute power to boost scientific discovery efforts and support our mission in advancing innovation in energy technologies,” said Steve Hammond, Director of NREL’s Computational Science Center. “By collaborating with HPE, we are gaining better tools to improve simulation and modelling across complex events to unlock new insights.”

Eagle will be installed in NREL’s Energy System Integration Facility (ESIF) data centre shortly and put into production use in January 2019.

Image credit: ©stock.adobe.com/au/CurvaBezier