Sign In

Communications of the ACM

ACM TechNews

Computing on Multiple Graphic Cards Accelerates Numerical Simulations

two-phase fluid flow

High performance fluid dynamics code is a goal of researchers at INS and Fraunhofer SCAI.

Credit: M. Griebel, P. Zaspel / INS, University of Bonn

The Fraunhofer Institute for Algorithms and Scientific Computing (SCAI) and the University of Bonn have been chosen as one of the first German Nvidia CUDA research centers. The researchers will focus on the development of paralleled multi-graphics processing units (GPUs) software for numerical simulation. "Our vision is to develop a massively parallel, completely multi-GPU-based high-performance molecular dynamics software package, as well as a massively parallel, completely multi-GPU-based high-performance fluid dynamics code," says SCAI professor Michael Griebel.

Numerical simulations can take days to compute, but SCAI's research could significantly shorten that time. The CUDA parallel computing architecture uses a GPU's massive computing power to increase computing performance.

The researchers want to modify the Tremolo-X software package, which is used for the molecular dynamics of atoms or molecules, for use on multiple graphics cards. Tremolo-X simulates materials at the nano scale, making it possible to efficiently design new and innovative materials. The GPUs also are much more energy efficient than standard CPUs.

From Fraunhofer SCAI
View Full Article


Abstracts Copyright © 2011 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account