AMBER Archive (2004)

Subject: RE: AMBER: Problem with sander tutorial.

From: Ross Walker (
Date: Thu Jul 22 2004 - 15:57:20 CDT

Dear Tarek,
I have observed this problem myself with older beta versions of the intel
fortran compiler 8.0. Can you download the very latest version from the
intel site and try this... If you still have problems email me back with the
details and the exact specs of your machine, and the exact version number of
the intel compiler you are using.
All the best
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- |
| | PGP Key available on request |


From: [] On Behalf Of
tarek mahfouz
Sent: 22 July 2004 10:00
Subject: RE: AMBER: Problem with sander tutorial.

Hi Ross,
Thanks for our response. I think we have a problem with the amber7 installed
on our unix system.
We compiled amber7 on a linux machine yesterday (gentoo linux) using intel
fortran compiler version 8 and we were trying to run the tests. Except for
different results for energy values from what you got, it run ok.
However, we had the following problem with the ubiquitin and dhfr tests. The
error message said "non bond list overflow". We decreased the cut off value
(from 9 to 8) but still the problem exists. Here is what the message said:

 using   5000.0 points per unit in tabled values
 TESTING RELATIVE ERROR over r ranging from 0.0 to cutoff
| CHECK switch(x): max rel err =   0.2738E-14   at   2.422500
| CHECK d/dx switch(x): max rel err =   0.8314E-11   at   2.736960
 * NB pairs          916     8243196 exceeds capacity (     8243681)   0
     SIZE OF NONBOND LIST =    8243681
 EWALD BOMB in subroutine ewald_list
 Non bond list overflow!
 check MAXPR in locmem.f

We noticed that the capacity changes as the size of non bond list changes so we always get this error. So, what would be the parameter to change here? Thanks, Tarek

----------------------------------------------------------------------- The AMBER Mail Reflector To post, send mail to To unsubscribe, send "unsubscribe amber" to