Re: [AMBER] combination of CHAMBER prmtop and pmemd.cuda is causing serious instability

From: Marc van der Kamp <marcvanderkamp.gmail.com>
Date: Wed, 11 Jan 2012 16:53:08 +0000

Hi Scot,

Just a quick reply through the mailing list, to make sure that my emails
directly to you haven't been classed as spam or so.
On 21/12 I sent you a tarball with the test case, as you requested.
Yesterday, I sent you a quick email to ask how things were going and if I
could be of any further help.
It would be useful to know roughly when a fix can be expected (weeks or
months?), to plan my current and future work. I fully appreciate that you
are busy and may have very little time trying to solve this problem.

Many thanks for all your excellent pmemd.cuda work!

Marc

On 21 December 2011 15:34, Scott Le Grand <varelse2005.gmail.com> wrote:

> Two requests:
>
> Marc, can you send me your test case?
>
> And Mark, could you zip up the pmemd/src/cuda directory resulting from the
> patch and send that to me?
>
> On Wed, Dec 21, 2011 at 4:29 AM, Mark Williamson <mjw.mjw.name> wrote:
>
> > On 12/21/11 10:07, Marc van der Kamp wrote:
> > > On 20 December 2011 19:23, Mark Williamson <mjw.mjw.name> wrote:
> > >
> > >> On 12/20/11 16:20, Marc van der Kamp wrote:
> > >>> On 20 December 2011 12:18, Mark Williamson <mjw.mjw.name> wrote:
> > >>>
> > >>>> On 12/20/11 10:22, Mark Williamson wrote:
> > >>>>> On 12/20/11 10:03, Marc van der Kamp wrote:
> > >>>>>> Hi Ross,
> > >>>>>>
> > >>>> Thanks, Mark.
> > >>> FYI, I was curious and did end up doing a complete fresh installation
> > of
> >
> > >> Hi again,
> > > Yes, the CA RMSD rises really, really quickly - as before. It goes
> from 1
> > > to 8 Ang in 50 ps.
> > > I expected as much (as the energies behaved pretty much the same as
> > > before), but I couldn't test a longer run until now.
> > > Thanks,
> > > Marc
> >
> >
> > Hi Marc,
> >
> > OK, thanks for the feedback and clarification about this; it is going to
> > need more investigation.
> >
> > For the moment, the ineffective "fix" for your issue will be pulled from
> > bugfix.20, but the other two fixes will remain since they do solve their
> > respective issues. I anticipate a bugfix.21 for your issue once a
> > workable fix is complete.
> >
> > Regards,
> >
> > Mark
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jan 11 2012 - 09:00:04 PST
Custom Search