Re: Help building a desktop for namd

From: Axel Kohlmeyer (
Date: Sun Apr 10 2011 - 12:52:59 CDT

On Sun, Apr 10, 2011 at 1:40 PM, Salvador H-V <> wrote:
> Hi all,
> I am new to the biomolecular world of computer simulations. I would like to
> learn from scratch how to use namd.
> I am thinking to build a basic desktop computer to run the tutorials and
> some simulations.
> At this moment, I think that phenom II x6 1000T amd  (six-core) processors
> are my best option (economic reasons). However, I was wondering if you could
> share some of your expecience using namd and vmd with amd processors.

if you want to learn, _any_ reasonable machine will do.
processors, compilers, nothing really makes much of a
difference unless you want to run as fast as possible,
but for that, you don't need a desktop, but a cluster.

thus, my advice. get a decent, well balanced, and reliable machine.
stay away from extremes (clock, memory, prices) in any direction.

> Besides, I am also worried about the performance of amd processors with free
> intel compilers. I know that free intel compilers perform very well on I7,i6
> & i3 but what about namd proccesors, any idea about this issue?

gcc also performs very well. the times that a compiler mad a huge
difference are long gone.

> Finally, I am thinking to use the six-core amd proccesors togehter with a
> GeForce GTX 580, to be able to use CUDA with namd and vmd, there is
> something that I should be aware of such combination?

you'd be getting much more bang for the buck,
if you'd get, say, two GT 560Ti cards.

the major problem with any of those is power and heat.

overall, for a desktop, the most important commodity is memory.
i'd rather double the memory and have less CPU or GPU power.

..and as i said before. any serious calculation will have to be
run on a serious machine anyways, so building a desktop for
that is a waste of energy and money.

> Any help or suggestion would be greatly appreciated.
> Thanks a lot,
> Salvador HV
> P.S I will use at least 8gb in ram

Dr. Axel Kohlmeyer
Institute for Computational Molecular Science
Temple University, Philadelphia PA, USA.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:56:57 CST