AMBER Archive (2000)

Subject: Problem with PME test

From: Jithesh P.V (jithesh_at_cdac.ernet.in)
Date: Thu Aug 10 2000 - 07:01:55 CDT

Hi,

I've compiled AMBER v 5.01 on a cluster of Sun Ultra 450 Workstations with
SunOS 5.6 on MPI. Since I intended to run PME simulations, the patch given
on the bugfixes page (bugfix.17 and bugfix.30 for ew_fft.f) were applied
before compilation with f77 compiler.

While running PME simulations with sander, the following error occured:

--------------------------
SANDER testing PME version:
 All processors started
 NUM_RECIP 1 NUM_DIRECT 3
 EWALD BOMB in subroutine grid_pointers (ewald.f):
 Cell neighbor pointer list too big
 Increase MAXNPTRS in ewald.inc
 EWALD BOMB in subroutine grid_pointers (ewald.f):
 Cell neighbor pointer list too big
 Increase MAXNPTRS in ewald.inc [1] MPI Abort by user Aborting program !
[1] Aborting program! [3] MPI Abort by user EWALD BOMB in subroutine
grid_pointers (ewald.f):
 Cell neighbor pointer list too big
 Increase MAXNPTRS in ewald.inc [2] MPI Abort by user Aborting program !
[2] Aborting program!
 Aborting program !
[3] Aborting program!
----------------------------

Could you please suggest a solution?

Thanks!

Jithesh
----------------------------------------
National PARAM Supercomputing Facility
Centre for Development of Advanced Computing
Pune, India
jithesh_at_jithesh.net
----------------------------------------