AMBER Archive (2001)

Subject: Re: periodic box size

From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Fri Feb 16 2001 - 14:35:36 CST

> In amber6, the box information from restart files will override A, B,
>and C in your input file. The box information specified in your
> input file will override that of parm file.
        
        Does this mean box sizes in the restrt file will _always_
        override those in prmtop, or only when ntx=6 or 7?

Only if the box is read (ntx=6 or 7).
        
>If you're using older version, I remember
>the box size should be better dividable by 2, 3 or 5
>to make the calculation faster.

This had to do with the ewald charge grid size (NFFTX,Y,Z in
http://www.amber.ucsf.edu/amber/doc/sander.txt), only loosely
related to the box dimensions.

>The larger the box size, more time is needed.

Better to say, the more atoms, the more time is needed.

Bill Ross