Hi all,
I'm getting a division by zero error when trying to run mmpbsa.pl on a
new 8-GPU host. The input/executables in question run perfectly fine on
another computer, so my guess is that the new server is missing
something in its configuration.
Both new and reference systems use the same executables and share the
same environment variables. All fortran libraries are accessible (Intel)
and MKL is visible. The system is a CentOS 5.5, kernel 2.6.18-194.el5.
pmemd.cuda ran with no issues.
Here's the error message:
[sasha.shasta cuda.7]$ /data/amber9/exe/mm_pbsa.pl
/scratch/cuda.7/mmpbsa/binding.in > /scratch/cuda.7/mmpbsa/energy.log
forrtl: severe (71): integer divide by zero
Image PC Routine Line
Source
sander 000000000092D060 Unknown Unknown Unknown
sander 0000000000927E31 Unknown Unknown Unknown
sander 00000000004457E2 Unknown Unknown Unknown
sander 00000000006BBD12 Unknown Unknown Unknown
sander 00000000004B91B8 Unknown Unknown Unknown
sander 00000000004AA77E Unknown Unknown Unknown
sander 00000000004A6700 Unknown Unknown Unknown
sander 00000000004002FE Unknown Unknown Unknown
sander 0000000000EBA3B0 Unknown Unknown Unknown
sander 00000000004001E9 Unknown Unknown Unknown
/data/amber9/exe/sander -O -i sander_com.in -o sander_com.1.out
-c /scratch/cuda.7/mmpbsa/snapshot_com.crd.1 -p
/scratch/cuda.7/complex.prmtop not successful
No other errors are logged. Can anyone suggest a possible source of this
error?
Thanks in advance
Sasha
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Aug 19 2010 - 12:00:04 PDT
