[CP2K:3926] problem when printing MO eigenvalues

hut... at pci.uzh.ch hut... at pci.uzh.ch
Tue Jul 17 15:32:16 UTC 2012


these output options (matrices and MO coefficients) to human readable
files were only intended for small systems. It works by 
collecting all the information on a single core and printing out
the data from this core. For systems of the size you are
looking at, this amounts to huge arrays and I can easily see
that you run into memory problems.
I don't know what data you really need, but the MO coefficients,
for example, can be found in the DFT restart file.
If you need more info, I suggest to write a specific module.



Juerg Hutter                         Phone : ++41 44 635 4491
Physical Chemistry Institute   FAX   : ++41 44 635 6838
University of Zurich               E-mail:  hut... at pci.uzh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland

-----cp... at googlegroups.com wrote: -----
To: cp... at googlegroups.com
From: Wei 
Sent by: cp... at googlegroups.com
Date: 07/17/2012 12:22AM
Subject: [CP2K:3926] problem when printing MO eigenvalues

Dear all,

Recently, we encountered a problem (probably caused by memory) when performing a non self-consistent calculation of a big system (more than 3,500 atoms). CP2K printed out the overlap matrix (AO*Log file, 67 GB) smoothly but always aborted when printing the MO eigenvectors & eigenvalues (MO*Log files).

By increasing the amount of memory per process up to 24 GB per process, a smaller system (2000atoms) worked out, but the 3500 atoms still fails even with up to 256 GB memory/process. Since we have performed the geometry optimization for this big system with quite some efforts, the more important thing for us is to calculate some physical property based MO eigenvalues. 

So I wonder if any of you have met this kind of problem or do you have any suggestion about it. Any help will be highly appreciated! Thanks a lot in advance! 

PS: I have tried it with both an old stable version and the most recent one from the trunck.

Best Wishes

 You received this message because you are subscribed to the Google Groups "cp2k" group.
 To view this discussion on the web visit https://groups.google.com/d/msg/cp2k/-/RSBWL3owf-IJ.
  To post to this group, send email to cp... at googlegroups.com.
 To unsubscribe from this group, send email to cp2k+uns... at googlegroups.com.
  For more options, visit this group at http://groups.google.com/group/cp2k?hl=en.

More information about the CP2K-user mailing list