Dear Amber users,
i have a strange problem with memory allocation in ptraj (amber7) on an
opteron with 16 GB of RAM running fedora core 3.
I am able to do a hbond analysis of 3075 frames but not 3100 frames or
more although there should be enough free memory.
With 3075 sets:
INPUT COORDINATE FILES
File (/home/fbo/md2500.crd.gz) is an AMBER trajectory (with box info)
with 2000 sets
File (/home/fbo/md3000.crd.gz) is an AMBER trajectory (with box info)
with 1075 sets
ACTIONS
HBOND saved to series hbonds, output to file /home/fbo/hbonds.data,
data will be sorted, intra-residue interactions will NOT be included,
Distance cutoff is 3.00 angstroms, angle cutoff is 120.00 degrees
Hydrogen bond information will be dumped for occupancies > 0.00
Estimated memory usage for this hbond call: 6109.21 MB
donors: 376 acceptors: 448
solvent donors: 7942 acceptors 15884
Processing AMBER trajectory file /home/fbo/md2500.crd.gz
Set 1 ..
But with some more frames (3100):
INPUT COORDINATE FILES
File (/home/fbo/md2500.crd.gz) is an AMBER trajectory (with box info)
with 2000 sets
File (/home/fbo/md3000.crd.gz) is an AMBER trajectory (with box info)
with 1100 sets
ACTIONS
HBOND saved to series hbonds, output to file /home/fbo/hbonds.data,
data will be sorted, intra-residue interactions will NOT be included,
Distance cutoff is 3.00 angstroms, angle cutoff is 120.00 degrees
Hydrogen bond information will be dumped for occupancies > 0.00
Estimated memory usage for this hbond call: 6158.83 MB
donors: 376 acceptors: 448
solvent donors: 7942 acceptors 15884
Processing AMBER trajectory file /home/fbo/md2500.crd.gz
ERROR in safe_malloc: Error in alloc of -2144460096 bytes
As said before, there are 16 GB available and other programs, like a big
gnuplot job, are able to use 12 GB without a problem.
If you need more information i can provide input files, compiler
versions etc.
Bye,
Fabian
--
Fabian Bös
Institute of Technical Biochemistry
University of Stuttgart / Germany
Phone: +49-711-6857481
Fax: +49-711-6853196
Email: fabian.boes.itb.uni-stuttgart.de
http://www.itb.uni-stuttgart.de
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Mon Dec 13 2004 - 12:53:01 PST