Re: [AMBER] error 12-6-4 LJ with the last update (November 2017)

From: Jason Swails <>
Date: Thu, 18 Jan 2018 13:49:44 -0500

On Thu, Jan 11, 2018 at 8:49 AM, Davide Sala <> wrote:

> ntb is > 0. I used the same protocol several times wothout any problem,

‚ÄčThen you were running without the r^-4 term, which is not what you thought
you were running. The code to evaluate the r^-4 term in the 12-6-4
potential isn't in pmemd.cuda anywhere (it was not "removed" by update.10),
so it's impossible to have been *actually* using it. I'd recheck the
previous results you ran with pmemd.cuda and the 12-6-4 potential... The
only CUDA implementation of 12-6-4 I'm aware of at the moment is OpenMM.

It's possible I'm missing something here (I thought there was an error
catch for it when Amber 16 was released, but I may be mistaken).

All the best,

Jason M. Swails
AMBER mailing list
Received on Thu Jan 18 2018 - 11:00:02 PST
Custom Search