AMBER Archive (2007)Subject: Re: AMBER: What is the best computer for running AMBER simulations?
From: Prem Prakash Pathak (prempolymer_at_gmail.com)
Date: Sat Aug 18 2007 - 04:36:00 CDT
Dear Amber user,
I am working solution structure of a 20kDa protein.
I want to run a MD simulation.
I have done a 100ps MD using sander, i.e. nstlim=50000 and dt=0.002,
and ntpr=50 in explicite water system with charge neutralized.
i wanted to know if these many steps are sufficient for a 20Kda
protein to get it acceptable.
regards
prem
On 8/17/07, Lili Peng <lilipeng_at_gmail.com> wrote:
> Thanks everyone for your input. This predicament I will ponder for a few
> days and make my final decision.
>
> On 8/16/07, Yu Chen <chen_at_hhmi.umbc.edu> wrote:
> >
> > You definitly don't want a MacBook Pro with Windows XP as main OS, that's
> > just waste of money and computing power.
> >
> > If you really are going to run computation intensive Amber jobs, then get
> > a 8-core Mac Pro desktop, much better for the job. Although if you only
> > run lots of small Amber jobs, then Macbook Pro will be fine, and will give
> > you mobility.
> >
> > CY
> >
> > >>> 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.
> > >
> > > This seems an odd combination to me. Using Windows for Amber can be a
> > pain,
> > > especially for long-running simulations. Using the Mac OSX should be
> > fine for
> > > Amber, and obviously is known to work well with Apple hardware. See
> > > Mengjeui's nice blog about Amber and Mac's on the Amber web site.
> > >
> > > I've found it hard to set up parallel Amber for windows (it's not really
> > > supported, but is do-able), and long-running processes can sometimes get
> > > interrupted. Please read carefully the section of the Amber web page
> > about
> > > Running Amber on MS Windows (recently updated). If your primary purpose
> > is to
> > > run compute-intensive calculations, I would find it hard to recommend
> > windows
> > > (unless you are a windows guru.)
> > >
> > > I would also second what Andreas said: consider getting a desktop
> > machine
> > > (probably with multiple cores and Linux) for the compute-intensive
> > stuff, plus
> > > a separate laptop (if you need a laptop for other reasons):
> > >
> > >>
> > >> For the price of the above machine, you could get a decent quad core
> > >> desktop machine (core 2 duo at 2.66 GHz) plus a laptop, which is still
> > >> fast (for everyday work and all but the newest computer games) and
> > still
> > >> have some money left.
> > >
> > > my $.02....dac
> > >
> > > -----------------------------------------------------------------------
> > > The AMBER Mail Reflector
> > > To post, send mail to amber_at_scripps.edu
> > > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> > >
> >
> >
> > ===========================================
> > Yu Chen
> > Howard Hughes Medical Institute
> > Chemistry Building, Rm 182
> > University of Maryland at Baltimore County
> > 1000 Hilltop Circle
> > Baltimore, MD 21250
> >
> > phone: (410)455-6347 (primary)
> > (410)455-2718 (secondary)
> > fax: (410)455-1174
> > email: chen_at_hhmi.umbc.edu
> > ===========================================
> > -----------------------------------------------------------------------
> > The AMBER Mail Reflector
> > To post, send mail to amber_at_scripps.edu
> > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> >
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
|