On Thu, Mar 1, 2012 at 1:43 PM, M. L. Dodson <mldodson.comcast.net> wrote:
>
> On Mar 1, 2012, at 12:18 PM, Jason Swails wrote:
>
> > On Thu, Mar 1, 2012 at 1:08 PM, Aron Broom <broomsday.gmail.com> wrote:
> >
> >> Dear AMBER Developers,
> >>
> >> I realize you might be getting tired of all the questions concerning the
> >> imminent release of AMBER 12, but I'm extremely interested to know if
> >> things like Adaptively Biased Molecular Dynamics will be made available
> in
> >> the cuda implementation in AMBER 12? I've been using it in sander, but
> the
> >> performance-to-cost boost of using AMBER on GPUs is so extreme at the
> >> moment when using something like a GTX570 in a desktop, that it is very
> >> hard to justify using sander.
> >>
> >
> > No, ABMD will not be making it into pmemd.cuda. (None of the NCSU codes
> > are being migrated over). Accelerated MD should be included, though, as
> an
> > alternative to ABMD.
> >
> > HTH,
> > Jason
> >
> >
>
> Since none of them will be available, what is the possibility of making the
> source code available under some open source license? This is quite often
> done with "end-of-life" code in commercial environments. If there will be
> some other alternative, then that might be sufficient, but it would be a
> shame
> for these very useful methods to just be rubbished. IMO.
>
They should still be available in sander (I haven't seen any activity in
the form of deep-sixing them), there's just no personnel/interest
combination working to port them to pmemd or the GPU version.
All the best,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Mar 01 2012 - 13:00:03 PST
