[gmx-users] Re: General conceptual question about advantage of GPUs

[Justin Lemkul]

On Tue, Apr 9, 2013 at 12:21 PM, Andrew DeYoung <adeyoung at andrew.cmu.edu>wrote:

> Thank you, Justin.  This is very helpful to me, especially the link to the
> GPU conference, which I think will be very helpful.
>
> I have done a little benchmarking on our ~7000 atom system, and in those
> tests the scaling was excellent -- almost linear when comparing 4, 8, 12,
> 16, 24 CPUs.  I am using particle decomposition, though; the results might
> be different with domain decomposition, I guess.
>
>
Yes, DD should be much faster than PD.

-Justin

-- 

========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540)
231-9080http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================