Re: [AMBER-Developers] [AMBER] another segfault with sander14 and TI/softcore

From: Jason Swails <jason.swails.gmail.com>
Date: Thu, 18 Dec 2014 07:37:28 -0500

> On Dec 18, 2014, at 5:26 AM, Hannes Loeffler <Hannes.Loeffler.stfc.ac.uk> wrote:
>
>> Do we need to keep supporting the old (multiple parmtop) way of doing it?
>> Would a rewrite simplify both code and docs? Plenty of big clusters
>> are still on AMBER12 or earlier, so bear in mind that changed made
>> today may not reach users until 2020 anyway.
>
> I'm not sure how much the availability of certain software versions on
> larger installation matters. Those who need and want new features will
> have their own license anyway and can freely run this wherever they
> like.
>
> Ditching the old multi-parmtop code seems attractive from the TI
> perspective (I do not know for what else it is used) because it adds an
> unnecessary layer of complexity for setup and of course it would make
> the docs simpler (they generally would profit from some
> clarifications!). But the problem, I think, is that pmemd/14 is not
> fully backward compatible. Pmemd doesn't support TI simulations in
> vacuum, idecomp != 0 is not supported, there may be others.

For my part, I would love for the pmemd TI implementation to be back-ported to sander. That it is substantially faster is, to me, of secondary importance, really. The main advantages that it brings is that the sander implementation hijacks the multisander functionality, meaning that it *can’t* be used for other tasks (specifically replica exchange). The pmemd implementation does not suffer from that limitation.

However, back-porting will be a very large job... one that I find unlikely to gain much traction given its availability in pmemd. So at present we’re stuck waiting for a brave soul to attack this problem while in the meantime continuing to support the double-topology approach (since the alternative is to cut the functionality out completely, which I’m not a fan of).

All the best,
Jason

--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Thu Dec 18 2014 - 05:00:02 PST
Custom Search