Maximizing the potential of exascale computing
March 15, 2017
RICHLAND, Wash. –
Advances in science require vast computational resources. Next generation supercomputers will help create a smart power grid and enable breakthroughs like reconstructing microbiomes, designing chemicals or materials at the molecular level, and modeling and simulating weather and data for climate research.
Breakthroughs and innovations in these and other fields will depend on exascale computing systems, being designed now to be at least 50 times faster than the nation's most powerful supercomputers in use today.
To be prepared to take full advantage of exascale systems, researchers are collaboratively participating in developing the software ecosystems, the hardware technology and a new generation of computational science applications, referred to as the co-design process.
The Department of Energy's Pacific Northwest National Laboratory has been chosen to lead an Exascale Computing Project co-design center focused on graph analytics. Researchers will develop methods and techniques for efficient implementation of key combinatorial algorithms chosen from the four areas: smart grids, computational biology, computational chemistry and climate science. The methods and techniques will be captured in a unified software framework called ExaGraph, that targets future extreme scale computing.
Lawrence Berkeley National Laboratory, Sandia National Laboratories and Purdue University are partners in the Exagraph: Combinatorial Methods for Enabling Exascale applications co-design center, led by Mahantesh Halappanavar a senior research scientist at PNNL.
Tags: Computational Science, Supercomputer