Fwd: Low global exclusion count on parallel tempering, not on MD

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Wed Aug 21 2013 - 11:27:44 CDT

I notice now that part of the message was not dispatched. That is, the
system is unstable,
(atoms moving too fast), however at the lowest temperature of parallel
tempering, corresponding to that of nicely running MD.

I must confess that the system was not equilibrated as thoroughly as usua,
I was in a hurry to check if parallel tempering starts.
However, with MD rmsd was constant and all values in namd plot were OK.

I am new to parallel tempering. Does it require exhaustive equilibration
beyond what I told above/

francesco pietra

---------- Forwarded message ----------
From: Francesco Pietra <chiendarret_at_gmail.com>
Date: Wed, Aug 21, 2013 at 5:45 PM
Subject: namd-l: Low global exclusion count on parallel tempering, not on MD
To: NAMD <namd-l_at_ks.uiuc.edu>

On submitting a parallel tempering job (namd2.9) the error arises:

------------- Processor 0 Exiting: Called CmiAbort ------------
Reason: FATAL ERROR: Low global exclusion count! (721771 vs 721816)
System unstable or pairlistdist or cutoff too small.

cutoff 12.0
switching on
switchdist 10.0
pairlistdist 13.5 ;# cutoff +3.5

while, on the same machine (BlueGene/Q), MD plays well with the same
(standard) settings.

Advice would be greatly appreciated

francesco pietra

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:33 CST