Communications of the ACM
More Than a Game
Researchers say they have used graphics processing units to crunch certain data-heavy atrophysics calculations more than 100 times faster than conventional methods.
Credit: Eugene Computer Geeks
Researchers at the SLAC National Accelerator Laboratory and Stanford University say they have used graphics processing units (GPUs) to demonstrate a way to crunch certain data-heavy astrophysics calculations more than 100 times faster than conventional methods.
The researchers note that GPU computing has been used in many applications, including medical imaging, bioinformatics, computational chemistry, and fluid dynamics.
"To start with, we just wanted to understand what the GPUs could do and how to interface with them," says Siena College professor Matt Bellis.
The researchers taught themselves the CUDA computer language and studied two different problems. One examined a collection of galaxies and determined how far apart they were from one another. "With GPUs we can do the full calculation faster and with much less uncertainty," says Stanford researcher Debbie Bard. The second calculation involved analyzing the shapes of galaxies to determine how much they have been affected by gravitational lensing.
"It's like having supercomputer power without having to use a supercomputer," Bard says. She believes video-card computing holds even more potential for accelerating data processing--and thus the speed of scientific discovery.
From Symmetry Magazine
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found