[gmx-users] Desktop computer for simulations and visualisation

[David van der Spoel]

On Fri, 2003-05-16 at 02:38, Dallas Warren wrote:
> Was just looking for people's opinion on an idea I have.
> 
> Currently I use an old O2 for visualisation of my trajectories etc. using 
> VMD.  With the number of atoms that I am getting to now, that is getting to 
> be a bit too much for it.  So I am looking to get something that can 
> display things much more smoothly and faster.  Additionally, as all MD 
> people are inclined, more CPU time to use would be nice.
> 
> So the idea is to get one of the current fastest dual (or quad?) processor 
> PC, run Linux, then use it for visualisation as well as running some 
> simulations.  I was thinking something like Xeon or Athlon (have to look at 
> the extra speed etc for the extra cost), fastest card that VMD can work 
> with etc.
> 
> Are there any problems with that, from a technical or implementation view 
> point?  Or is this something that some people already do?
> 
> Of course there is all the issues to do with actual components, which one, 
> expense versus performance etc, but that is the next step :-)
> 
> I have noticed from discussions on this list that most people run the 
> Athlon in their clusters.  From the sound of it you get good bang for your 
> buck.  I also know that the couple of super computers I have CPU time on 
> down here use either Xeon or Alphaservers.  What about the Opteron (AMD)?
> 
> Ahh, the joy and confusion of purchasing a computer ;-)

Check the gromacs benchmarks for "objective" information. I have lately
been buying single P4 machines with decent graphics cards (even have a
stereo setup with Linux/NVidia, but mainly for the crystallographers
over here). The nice thing about these boxes is that you can get
relatively quiet (queiter than my laptop...). 
For computing I have latetely bought dual Xeons, and this is what we use
in the computer centre too. It would be nice to see some benchmarks on
the Opteron, in particular for double precision SSE2 ...

-- 
Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Dept. of Cell & Mol. Biology
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
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++