[gmx-developers] Using graphics coprocessors for MD

David spoel at xray.bmc.uu.se
Sat May 15 13:14:35 CEST 2004


On Mon, 2004-04-26 at 09:06, Lieven Buts wrote:
> On Friday 23 April 2004 11:15, Erik Lindahl wrote:
> > 4. The theoretical performance is incredibly (200Gflop), but it drops
> > very fast when you start to use more registers (complicated to explain,
> > but confirmed by ATI/NVidia). Our current performance is about 50% of
> > what you get from SSE on a recent P4.
> 
> While on the subject of speed: I noticed that the molecular modelling program 
> YASARA offers a 100 dollar prize for code that runs a specified simulation 
> faster than their code (benchmark 2):
> 
> http://www.yasara.com/benchmarks.htm
> 
> Surely GROMACS cannot let such a challenge go unanswered ;-) ?

Found back this mail and did the benchmark. GROMACS is exactly twice as
fast, 0.34 s / step compared to their 0.7 s / step. I mailed the results
to them.

-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++




More information about the gromacs.org_gmx-developers mailing list