[gmx-developers] Gromacs on Mare Nostrum (Barcelona Supercomputing Centre)
lindahl at cbr.su.se
Mon Apr 16 11:42:25 CEST 2007
We'll get in touch with them, but considering we've run tons of
protein in parallel without unfolding even with v 3.2.1, I think it's
more probably that this is an error with the compiler/setup/user.
We can't rule out a bug of course (in particular since PPC970 + IBM
compilers isn't such a common combination), but we're a bit irritated
that they don't even seem to have bothered getting in touch with is
to get help before claiming it's a bug...
On Apr 16, 2007, at 9:43 AM, Tsjerk Wassenaar wrote:
> Dear developers,
> Here is a link to a technical report comparing three different MD
> suites, including Gromacs (3.2.1/ffG43a1/PME), on the Barcelona
> supercomputer. I'd appreciate if one of you could give it a glance and
> a reaction. They mention consistent unfolding of proteins when using
> multiple processors.
> In the report they mention additional data on-site, but I haven't
> found it yet.
> I'll put in a request to check 3.3, but would like some additional
> input. There may also be added value in contact between the Gromacs
> crew and the Mare Nostrum one, but that may be considered later. Note
> that these efforts are done within the framework of DEISA and that the
> outcomes may be of some significance for the use of Gromacs.
> Tsjerk A. Wassenaar, Ph.D.
> Junior UD (post-doc)
> Biomolecular NMR, Bijvoet Center
> Utrecht University
> Padualaan 8
> 3584 CH Utrecht
> The Netherlands
> P: +31-30-2539931
> F: +31-30-2537623
> gmx-developers mailing list
> gmx-developers at gromacs.org
> Please don't post (un)subscribe requests to the list. Use the www
> interface or send it to gmx-developers-request at gromacs.org.
More information about the gromacs.org_gmx-developers