From: Jan Janssen (janssen_at_mpie.de)
Date: Thu Nov 10 2016 - 07:30:38 CST

Ok, just a quick update. We doubled the memory and everything worked fine. Thanks again for all the support. 

With best regards 

Jan 

 From: Jan Janssen <janssen_at_mpie.de>
 To: John Stone <johns_at_ks.uiuc.edu>
 Cc: Axel Kohlmeyer <akohlmey_at_gmail.com>, <vmd-l_at_ks.uiuc.edu>
 Sent: 11/10/2016 9:51 AM
 Subject: Re: vmd-l: Limits of the VMD Visualisation

 
 Hello John, Hello Axel, 

Turning off the Autobonds helped, I can now load files up to 50398638 atoms, but I get the following error message: 

ERROR) Failed to increase display list memory pool size, system out of memory

ERROR)   Previous pool size: 1068MB

ERROR)   Requested pool size: 1974MB

An when I further increase the sample size to 60398640 atoms, the system crashed already while reading the file:

Info) Using plugin xyz for structure file /home/cm/Downloads/pureAl_pos.xyz

Info) Using plugin xyz for coordinates from file /home/cm/Downloads/pureAl_pos.xyz

Info) Analyzing structure ...

Info)    Atoms: 60398640

Info)    Bonds: 0

Info)    Angles: 0  Dihedrals: 0  Impropers: 0  Cross-terms: 0                          

Info)    Bondtypes: 0  Angletypes: 0  Dihedraltypes: 0  Impropertypes: 0                

Info)    Residues: 60398640                                                             

Info)    Waters: 0                                                                      

Info)    Segments: 1                                                                    

Killed   

So I guess now it is an memory issue and the 16GB of system memory are simply not enough. I will try to get a 32GB system. 
Still it would be interesting if you have any further recommendations in terms of options or additional libraries to include during compilation to optimise handling such large files. 

Regards 

Jan 

 From: John Stone <johns_at_ks.uiuc.edu>
 To: Jan Janssen <janssen_at_mpie.de>
 Cc: Axel Kohlmeyer <akohlmey_at_gmail.com>, <vmd-l_at_ks.uiuc.edu>
 Sent: 11/10/2016 8:45 AM
 Subject: Re: vmd-l: Limits of the VMD Visualisation

Hi,
  Try running this command on the text interface in VMD:
mol new yourfile.xyz autobonds off
 
Cheers,
  John Stone
  vmd_at_ks.uiuc.edu
 
On Thu, Nov 10, 2016 at 08:29:50AM +0100, Jan Janssen wrote:
>    Hello Axel,
>
>    Thanks for the quick reply. I tried the Linux Version and it worked better
>    in terms that is was able to read in the file. It works fine for 2398638
>    Atoms, but even at this point I already get the error message:
>
>      Info) Determining bond structure from distance search ...
>      ERROR) MolAtom 1460117: Exceeded maximum number of bonds (12).
>
>    When I then try to increase the size even further, to something like
>    6398642 atoms, the process crashes:
>
>      Info) Determining bond structure from distance search ...
>      vmd_gridsearch_bonds (thread 4):  25.69 % complete
>      vmd_gridsearch_bonds (thread 0):  31.25 % complete
>      vmd_gridsearch_bonds (thread 8):  26.39 % complete
>      terminate called after throwing an instance of 'std::bad_alloc'
>        what():  std::bad_alloc
>      Abort (core dumped)
>
>    So I was wondering if there is an option to disable the "Bond structure
>    determination" especially for this kind of large samples?
>    Regards
>
>    Jan
>
>    From: Axel Kohlmeyer <akohlmey_at_gmail.com>
>    To: Jan Janssen <janssen_at_mpie.de>
>    Cc: Vmd l <vmd-l_at_ks.uiuc.edu>
>    Sent: 11/9/2016 7:21 PM
>    Subject: Re: vmd-l: Limits of the VMD Visualisation
>
>      On Wed, Nov 9, 2016 at 11:59 AM, Jan Janssen <janssen_at_mpie.de> wrote:
>      > Hello,
>      >
>      > I guess this is not exactly the purpose of VMD, but we try to
>      visualise an
>      > Atom Probe Dataset which we converted to *.xyz Format. The dataset
>      consists
>      > of 6 million atoms and at least the Mac OS X Version of VMD crashed
>      when
>      > trying to load the data, while loading a subset of the file worked
>      fine.
>      > The question is whether we can compile VMD with special packages to
>      support
>      > such large datasets or whether other file formats are more suitable
>      for
>      > loading such large datasets. The file size is currently about 2GB.
>
>      current downloads of VMD for MacOSX are 32-bit (regardless of the CPU
>      and OS) and thus you are limited in your address space.
>
>      it should work on linux with a 64-bit binary.
>
>      axel.
>
>      >
>      > Regards
>      >
>      > Jan
>      >
>      >
>      >
>      >
>      > Crash:
>      > Info) Using plugin xyz for structure file
>      /Users/jan/Desktop/pureAl_pos.xyz
>      > /Applications/VMD 1.9.3.app/Contents/MacOS/startup.command: line 7:
>      14731
>      > Segmentation fault: 11  "$p/../Resources/VMD.app/Contents/MacOS/VMD"
>      $*
>      > logout
>      > Saving session...
>      > ...copying shared history...
>      > ...saving history...truncating history files...
>      > ...completed.
>      >
>      >
>      > Head of the file:
>      > 60398640
>      > Atom probe data
>      > Al -8.14419269562 -44.8013191223 -14.9069433212
>      > Al -35.179271698 -3.4159450531 -8.59115982056
>      > Al -15.6410989761 -25.2453269958 -5.95746135712
>      > Al 1.84774339199 -44.2051582336 -13.9800004959
>      > Al 6.15377855301 -44.0587081909 -14.1510372162
>      > Al 17.3810424805 -50.0091285706 -21.0890655518
>      > Al -22.3692836761 -40.944026947 -15.7445049286
>      > Al -0.755965948105 -44.747303009 -14.3411016464
>      >
>      >
>      >
>      >
>
>      --
>      Dr. Axel Kohlmeyer  akohlmey_at_gmail.com  http://goo.gl/1wk0
>      College of Science & Technology, Temple University, Philadelphia PA, USA
>      International Centre for Theoretical Physics, Trieste. Italy.
>

--
NIH Center for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
http://www.ks.uiuc.edu/~johns/           Phone: 217-244-3349
http://www.ks.uiuc.edu/Research/vmd/      
 

-------------------------------------------------
Max-Planck-Institut für Eisenforschung GmbH
Max-Planck-Straße 1
D-40237 Düsseldorf
 
Handelsregister B 2533
Amtsgericht Düsseldorf
 
Geschäftsführung
Prof. Dr. Gerhard Dehm
Prof. Dr. Jörg Neugebauer
Prof. Dr. Dierk Raabe
Dr. Kai de Weldige
 
Ust.-Id.-Nr.: DE 11 93 58 514
Steuernummer: 105 5891 1000

Please consider that invitations and e-mails of our institute are
only valid if they end with …@mpie.de.
If you are not sure of the validity please contact rco_at_mpie.de

Bitte beachten Sie, dass Einladungen zu Veranstaltungen und E-Mails
aus unserem Haus nur mit der Endung …@mpie.de gültig sind.
In Zweifelsfällen wenden Sie sich bitte an rco_at_mpie.de
-------------------------------------------------