Actually, I have no problem with k-points parallelization in case of scf calculations.<br><br>The problem is when running *.<a href="http://ph.in">ph.in</a> using ph.x ( with elph=.true.,)- in this case I have to set npool =1 so the calculations will continue-.<br>
<br>I will try the Al example in the new version and let you know. Actually I was running this example ( example07 in espresso example directory) using espresso-4.0.3 and I had the same problem of using more than pool.<br>
<br>Many Thanks<br>IYAD<br><br><br><br><br><br><br><div class="gmail_quote">On Wed, May 16, 2012 at 3:51 PM, Paolo Giannozzi <span dir="ltr"><<a href="mailto:giannozz@democritos.it" target="_blank">giannozz@democritos.it</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im"><br>
On May 16, 2012, at 17:10 , Iyad AL-QASIR wrote:<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
What make things slow is I can not use npool >1. since this is not<br>
implemented yet for el-ph calculations<br>
</blockquote>
<br></div>
I vaguely remember that k-point parallelization was implemented<br>
some time ago. In any case, if you run the example for the<br>
electron-phonon case (PHonon/examples/example03 in the<br>
new directory structure) it seems to work with -npool 2<br>
<br>
P.<br>
---<br>
Paolo Giannozzi, Dept of Chemistry&Physics&Environment,<br>
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy<br>
Phone <a href="tel:%2B39-0432-558216" value="+390432558216" target="_blank">+39-0432-558216</a>, fax <a href="tel:%2B39-0432-558222" value="+390432558222" target="_blank">+39-0432-558222</a><br>
<br>
<br>
<br>
<br>
</blockquote></div><br><br clear="all"><br>-<br>