Earth Science Project To Consume 83 Million Processor Hours
- By Dian Schaffhauser
- 11/18/14
Scientists from seven institutions will be consuming 83 million processor hours from a supercomputer at Argonne National Laboratory to better understand why astronomical bodies, including the Earth, generate self-sustaining, large-scale magnetic fields. The project was made possible through an award program started a decade ago by the United States Department of Energy. INCITE — Innovation and Novel Computational Impact on Theory and Experiment — grants time on two of the largest supercomputers run by the Department, one at Argonne in Illinois and the other at Oak Ridge in Tennessee.
The planetary and stellar magnetism project is pulling together researchers from the Universities of California at Los Angeles, Davis, Berkeley and Santa Cruz, as well as the University of Colorado, Boulder, Johns Hopkins University and the University of Toronto. The team will use Argonne's "Mira" supercomputer to simulate the formation and progression of Earth's magnetic field using source code developed at UC Davis. The work will also involve simulating the magnetic field generation occurring inside Jupiter and the sun.
Data will be made freely available to the broader scientific community, Project Lead Jonathan Aurnou, a UCLA professor of planetary sciences and geo project leader said in a statement.
"We have never been able to simulate realistic magnetic field generation in turbulent liquid metal and plasma flows, even though that is what exists inside planets and stars," Aurnou said. "But this INCITE project will allow us, for the first time, to do just that."
This initiative was one of 56 announced last year by the Department of Energy. In total researchers have gained access to 5.8 billion core-hours on the two supercomputers.
Mira is a 10-petaflop, 786,000 core IBM Blue Gene/Q system capable of 10 quadrillion calculations per second. (That's equivalent to a million billions.) Mira was ranked fifth by the TOP 500 in a list of the world's fastest supercomputers. Titan, at Oak Ridge, is ranked second.
The award also includes support and assistance by INCITE staff to help researchers optimize their processing time on the computers.
About the Author
Dian Schaffhauser is a former senior contributing editor for 1105 Media's education publications THE Journal, Campus Technology and Spaces4Learning.