<div class="gmail_quote">On Sun, Jul 22, 2012 at 10:55 PM, sylvian <span dir="ltr"><<a href="mailto:sylviankahane@gmail.com" target="_blank">sylviankahane@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="white" background="cid:image001.jpg@01CD6863.63C2E360" lang="EN-US" link="blue" vlink="purple"><img width="0" height="0" style="width:0;min-height:0"><div><p class="MsoNormal"><span style="font-family:Arial,sans-serif;font-size:11pt">cp.x is printing an Etot at the end of the run (in my case something like -47 Ha). Taking its state at the end of the run, i.e. ions positions and charge density, as an input to pw.x (admittedly with a different pseudopotential) I am getting a "total energy" differing by an order of magnitude (in my case -800 Ry).</span></p>
<p class="MsoNormal"><br></p></div></div></blockquote><div><br></div><div><br></div><div>Dear Sylvan,</div><div>absolute energies have no meaning in general. There can be some more or less natural choices for a zero energy, but n periodic boundary conditions and with pseudopotential there is none that really makes sense.</div>
<div><br></div><div>It is also true that with fairly similar pseudopotential you often get similar (if equally meaningless) "total" energies. You are not specifying which pp's you're using so I can only guess:</div>
<div>1. the pp used in pw.x has more valence electrons</div><div>2. the pp used in pw.x is actually a paw dataset, and you are comparing the all-electron energy with the valence one</div><div><br></div><div>best regards</div>
</div><div><br></div>-- <br><div>Lorenzo Paulatto IdR @ IMPMC/CNRS & Université Paris 6</div><div>phone: +33 (0)1 44275 084 / skype: paulatz</div><div>www: <a href="http://www-int.impmc.upmc.fr/~paulatto/" target="_blank">http://www-int.impmc.upmc.fr/~paulatto/</a></div>
<div>mail: 23-24/4é16 Boîte courrier 115, 4 place Jussieu 75252 Paris Cédex 05</div><br>