<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Alan,<br>
I don't think using single precision is much of a problem when using
thermostats, regardless of the constraint on the water.<br>
<br>
See also Berks' comments:<br>
<a class="moz-txt-link-freetext" href="http://oldwww.gromacs.org/pipermail/gmx-users/2010-March/049137.html">http://oldwww.gromacs.org/pipermail/gmx-users/2010-March/049137.html</a><br>
<a class="moz-txt-link-freetext" href="http://oldwww.gromacs.org/pipermail/gmx-users/2010-March/049152.html">http://oldwww.gromacs.org/pipermail/gmx-users/2010-March/049152.html</a><br>
<br>
Ran.<br>
<br>
Alan wrote:
<blockquote
cite="midAANLkTinM3nxpfZaz4oPUPTNaRr2LeLOh3a6XqJHJWfg9@mail.gmail.com"
type="cite">
<div class="gmail_quote"><span class="Apple-style-span"
style="font-family: arial,sans-serif; font-size: 13px; border-collapse: collapse;">Hi
there,
<div><br>
</div>
<div>From what I've read and known, here in the list as well, one of
the main reasons why Gromacs run in single precision is because it has
LINCS, besides SHAKE, which I believe requires double precision for
accuracy.</div>
<div><br>
</div>
<div>I am drawing such conclusion (that can be wrong) partially based
on </div>
<div><br>
</div>
<div>Lippert, R. A., Bowers, K. J., Dror, R. O., Eastwood, M. P.,
Gregersen, B. A., Klepeis, J. L., Kolossvary, I., and Shaw, D. E. A
common, avoidable source of error in molecular dynamics integrators.
Journal of Chemical Physics 126, 4 (Jan. 2007), 046101–1–046101–2</div>
<div><br>
</div>
<div>However, in this letter article they didn't test with LINCS. I
would love to hear some comments from Gromacs developers.</div>
<div><br>
</div>
<div>When I started in MD, developing our own MD software, all was
done in double precision, then came Gromacs blowing up this paradigm.
(I am aware that even in Gromacs, there are routines that really
requires double precision, e.g. normal mode analysis).</div>
<div><br>
</div>
<div>Essentially I would like to understand better this double x
single approach in MD re accuracy.</div>
</span>
<div>
<div>
<div class="h5"><br>
<div class="gmail_quote">Thanks,</div>
<div class="gmail_quote"><br>
</div>
<div class="gmail_quote">Alan</div>
<div class="gmail_quote"><br>
</div>
</div>
</div>
</div>
</div>
-- <br>
Alan Wilter S. da Silva, D.Sc. - CCPN Research Associate<br>
Department of Biochemistry, University of Cambridge. <br>
80 Tennis Court Road, Cambridge CB2 1GA, UK.<br>
>><a href="http://www.bio.cam.ac.uk/%7Eawd28">http://www.bio.cam.ac.uk/~awd28</a><<<br>
</blockquote>
<br>
</body>
</html>