I am interested in watching the dynamical behavior for some material systems using DFT MD.  The time periods I am interested in range out to roughly 1 ns and I am experiencing problems with the front end running out of memory.  What options do I have to reduce memory usage -- I would, of course, like to keep track of the structure and how it is changing -- in particular I would like to calcluate periodically bonding configurations/lengths and ring statistics, but the need to keep track of the trajectory and/or temperature/pressure/energy variables every interation is not so accute (a disk file is fine).  How can I achieve this.  At the moment, the front end fails to even download results when the size is over a (still to be determined, but large) size.  Is the Materials Studio front end just not up to the task and I will have to parse everything manually?  Suggestions anyone?  For background a few 10s of thousand iteration steps leads to a log file in the GB range.  This is not really a problem and I want the data in the file, it just seems that MS is broken with respect to such large studies.

On another (related) subject, are there any standard scripts for printing out bonding configurations for a given structure?

Thanks,

             Paul Fons