From: Marc Q. Ma (qma_at_oak.njit.edu)
Date: Thu Feb 24 2005 - 16:52:19 CST
Hi Glenn,
Jim has explained clearly how these numbers should be derived. Two
factors: accuracy and efficiency. Basically, the larger the size, the
more computation. A rule of thumb is to choose the sizes that
correspond to 1 grid per 1 AA, which gives good accuracy. You can
choose to have 1 grid per 1.5 or 2 AA, however, the accuracy is not as
good. For best efficiency, you use sizes such as 30, 32, 36, 40, 45,
48, 50, 54, 60, 64, 72, 75, 80, 81, and 90. If the basis vector is
{45.8, 0, 0}, you may use 48, 45 or 40 for good accuracy and
efficiency. If you choose to use 36, 32 or 30, the computation will be
very fast, but the accuracy is not good.
Marc
On Feb 24, 2005, at 4:25 PM, Glenn Johnson wrote:
>
>
> I am trying to get an understanding of the PMEGridSize{X,Y,Z}
> parameters. I searched the mail archive and found the following
> response:
>
> http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/0051.html
>
> but I do not understand how those numbers were derived. Could someone
> please enlighten me?
>
> Thanks.
>
> --
> Glenn Johnson
>
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:40:32 CST