From: Chris Harrison (charris5_at_gmail.com)
Date: Tue Mar 27 2012 - 10:15:36 CDT
Francesco,
We're going to need more info. Since RAMD is predominately a Tcl-script
based implementation at present, your particular input, any altered
scripts, and underlying parallel infrastructure (charm++ version, mpi vs
non-mpi, network, etc) could be involved as a potential problem.
More detail on the output, or a copy of the last 20 lines of the log file
might begin to help.
Best,
Chris
On Tue, Mar 27, 2012 at 9:56 AM, Francesco Pietra <chiendarret_at_gmail.com>wrote:
> Hi:
> I deleted previous mail, where I posted that that the binary linux 64
> cuda night build 2012-03-23 had solved problems of crashing in
> min/RAMD, which I encountered with the standard 2.9b1. Actually, in my
> hands, there are still problems, at least on running RAMD for longer
> than 300,000 steps. Repeatedly, the simulation halted, going to the
> linux prompt, without any error or warning message. Something that I
> had never seen before.
>
> Back to 2.8, I have carried out two RAMD simulations - with the same
> files - for longer than 600,000 steps. Each one closed regularly
>
> System Debian amd 64, two GPU-580. FF parm7.
>
> francesco pietra
>
>
-- Chris Harrison, Ph.D. NIH Center for Macromolecular Modeling and Bioinformatics Theoretical and Computational Biophysics Group Beckman Institute for Advanced Science and Technology University of Illinois, 405 N. Mathews Ave., Urbana, IL 61801 http://www.ks.uiuc.edu/Research/namd Voice: 773-570-6078 http://www.ks.uiuc.edu/~char Fax: 217-244-6078
This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:21:48 CST