Hi Lili,
What kind of systems are you considering? If these are large-ish systems (>
50,000 atoms), I would think a laptop would not be an ideal choice for MD at
all because you will not be able to really benefit from the parallel scaling
abilities in SANDER, and especially PMEMD. However, for small systems of
less than a few thousand atoms, and for using non-MD modules like
leap/ptraj/etc, your laptop may not be such a bad idea.
As a linux-minded person, and after working with Windows and Mac's for
several years now, I would not recommend jeopardizing your Mac's stability
by allowing Windows to run anywhere near it. OSX is generally a stable
operating system because it has a unix backbone, but I don't think the same
level of stability can be found in any line of MS operating systems. So,
if I were in your shoes I would buy a Macbook with OSX, not with XP. That
said, I've had the most overall success with linux's stability (uptime of
nearly two years in some cases), and believe you can find installations for
the Intel's Core 2 chips for a variety of distributions. The benefit is
that you should be able to run AMBER without the need of Cygwin, as is the
case for Windows, but the drawback may be an intense initial learning curve
associated with first time linux use.
Just my opinions...
Good luck,
David LeBard
On 8/16/07, Lili Peng <lilipeng.gmail.com> wrote:
>
> Dear all,
>
> I am in the process of purchasing a new laptop for computational modeling
> with AMBER. After consulting with my peers and colleagues, I was encouraged
> to get a Macbook Pro with the Windows XP operating system installed on
> it. The following tech specs was suggested to me:
>
> 2.4GHz Intel Core 2 4MB L2 Cache
> 4GB DDR2 PC2-5300
> 7200 rpm
> 256 MB NVIDIA GeForce 8600M GT
> Superdrive 8x dual layer
> iSight
> Apple Remote
> Airport Extreme
> Bluetooth 2.0
> Firewire 400 & 800
> (5.4 lbs)
>
> I'd like to note that the most important factor in my purchasing a
> computer would be processing time, given that I will be running
> computationally-intensive simulations. Are these a good set of specs for
> this purpose? Another colleague of mine also suggested that I look into
> Lenovos (formerly ThinkPads), but I heard Macbooks are more stable.
>
> If anyone has had any experience with purchasing computers designed
> specifically for the fastest computation time, and would be willing to share
> your knowledge, I would be most appreciative of your insights.
>
> Thanks,
> Lili
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Aug 19 2007 - 06:07:31 PDT
