AMBER Archive (2005)

Subject: AMBER: problem with nmode using ntrun=5

From: Phineus Markwick (markwick_at_embl-heidelberg.de)
Date: Fri Mar 11 2005 - 08:38:34 CST


Dear AMBER users,

I have met a problem when calculating langevin modes in nmode (using
ntrun=5).
The program exits with the error message in the output file:

 expfile is not present. Hydrodynamic radius
  of 0.7700 angs. is used for all atoms
  except hydrogen, where max. value is 0.2

| Time to set up A matrix: 0.0000
** On entry to DGEEV parameter number 13 had an illegal value

I recognise that this is a problem with the lapack libraries, and not
AMBER as such.
I have searched the amber email reflector archives and bug-lists, and
although I have not found
this particular problem, bugfix 1 for AMBER7 deals with a problem with
the lapack libraries
aswell and more info is on the archive list (
http://amber.ch.ic.ac.uk/archive/200407/0061.html ).

I have had no problems as yet with the lapack libraries, and before
changing the Makefile, I was wondering
if anybody else has met this problem.

I am using AMBER7 on Linux 2.4 21-27.0.2ELsmp, x86_64 GNU/Linux,
2x2.2GHz AMD Opteron CPU's
and the optimisation level appears to be O6 using the GCC compiler for
Opteron.

If anybody has met this problem before, I would be grateful if they
could tell me if it really is just
an optimisation level problem for the Lapack libraries, or give me some
advice on how to proceed.

many thanks in advance for any suggestions,

with best regards,
Phineus Markwick.

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu