Re: Info: Pairlistdist is too small for 3 computes during timestep 115001

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Fri Dec 12 2014 - 15:49:06 CST

This is not an error. The pairlist distance adjusts on a per-patch basis,
and it for some timesteps it is too short before the next pairlist
recalculation the pairlist is discarded for that compute only.

It is strange that your pairlists are invalid after a single step, which
suggests some atoms are moving particularly fast, like a gas or otherwise
unstable.

The parameters that control the adjustment are:

   pairlistShrink (default 0.01), tol *= (1 - x) on regeneration

   pairlistGrow (default 0.01), tol *= (1 + x) on trigger

   pairlistTrigger (default 0.3), trigger is atom > (1 - x) * tol

Increasing pairlistTrigger would result in a higher pairlist tolerance.

Jim

On Fri, 12 Dec 2014, Santosh Kumar Chaudhary wrote:

> Dear All,
>
> I am new to NAMD simulations. While running the simulation i am getting an
> info "Info: Pairlistdist is too small for 3 computes during timestep
> 115001" after every outputpairlist, which was set to 1000. I have tried
> changing the
> cutoff 10 to 12,
> switchdist 8 to 10
> pairlistdist 12 to 14 but nothing is working. Simulation runs fine only
> the info appears after every 1000 steps. Please Help me to understand the
> problem and suggest the required solution.
>
> Thanks and regards
> Santosh Chaudhary
> Supercomputer Education and Research Center
> Indian Institute of Science
> Bangalore - 560012
>
>
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.
>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:28 CST