<div>What I meant is not to use any processor specific options when compiling PW using ifort. For example, "-xavx" for new Sandy bridge processors.</div><div> </div><div>From my (limited) experience I noticed that if you use "-O2" only, I never had any problems.</div><div> </div><div>Sergey</div><div> </div><div> </div><div>04.04.2013, 15:45, "TRINITE Virginie" <virginie.trinite@thalesgroup.com>:</div><blockquote type="cite"><div><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US">Thanks for your reply.</span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US">I am not sure that I understand what you mean. Deed you mean that you drop all flags of the compiler?</span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US">What compiler did you use?</span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US"> </span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US">Best Regards</span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US"> </span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US">Virginie Trinite</span></p><p><span style="font-size:11pt;font-family:Calibri,sans-serif;color:#1f497d;" lang="EN-US"> </span></p><div style="border:none;border-top:solid #B5C4DF 1pt;padding:3pt 0cm 0cm 0cm;"><p><strong><span style="font-size:10pt;font-family:Tahoma,sans-serif;">De :</span></strong><span style="font-size:10pt;font-family:Tahoma,sans-serif;"> <a href="mailto:pw_forum-bounces@pwscf.org">pw_forum-bounces@pwscf.org</a> [mailto:<a href="mailto:pw_forum-bounces@pwscf.org">pw_forum-bounces@pwscf.org</a>] <strong>De la part de</strong> Sergey Lisenkov<br /> <strong>Envoyé :</strong> vendredi 29 mars 2013 15:42<br /> <strong>À :</strong> PWSCF Forum<br /> <strong>Objet :</strong> Re: [Pw_forum] PW is frozen</span></p></div><p> </p><div><p>I had the same issue (the code stopped to write any output in the middle of calculations). I then recompiled the code without using any "-xXXX" options and it seems solved the problem.</p></div><div><p> </p></div><div><p>Sergey.</p></div><div><p> </p></div><div><p> </p></div><div><p>29.03.2013, 14:05, "TRINITE Virginie" <<a href="mailto:virginie.trinite@thalesgroup.com">virginie.trinite@thalesgroup.com</a>>:</p></div><blockquote style="margin-top:5pt;margin-bottom:5pt;"><div><p><span lang="EN-US">Dear all</span></p><p><span lang="EN-US"> </span></p><p><span lang="EN-US">I find a very strange behavior of pw.x:</span></p><p><span lang="EN-US">                It run endlessly without output anything and with no error, and I have  finally to kill it</span></p><p><span lang="EN-US">This behavior appears by using 6 processors for Iron calculation, everything is fine otherwise (and I made number of other calculations with no error with this version).</span></p><p><span lang="EN-US">I use espresso-5.0 with ifort 12.1.3 and openmpi-1.4.3i</span></p><p><span lang="EN-US">I find a very old post on this type of problem: <a href="http://www.democritos.it/pipermail/pw_forum/2004-December/001733.html">http://www.democritos.it/pipermail/pw_forum/2004-December/001733.html</a>.</span></p><p><span lang="EN-US">It seems related to vc-relax so I try scf calculation but it was the same</span></p><div><div><p><span lang="EN-US">I try to run directly on the node without the queue system and it was the same: no output on the screen for 6 processors and 1 2 3 4 5 processors is fine.</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">I try to clean all the files, in case pw.x try to read old wavefunctions, without solving the problem.</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">In the old post Paolo Giannozzi explain that can be due to MPI or MKL library. But since 2004, I think that the libraries has a lot evolved?</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> </span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">Did anybody have this kind of behavior?</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> </span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">Here is the input file:</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> </span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">&control</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    calculation   = 'scf'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    verbosity     ='high'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    outdir        = '/tmp/'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    prefix        = 'feFM'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    tstress       =.true.</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    tprnfor       =.true.</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    forc_conv_thr = 1e-4</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> /</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> &system</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    ibrav         = 2</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    celldm(1)     = 6.524</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    nat           = 1</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    ntyp          = 1</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    ecutwfc       = 120</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    ecutrho       = 1200</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    starting_magnetization(1)=1.0</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    occupations   = 'smearing'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    smearing      = 'mp'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    degauss       = 0.01</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    nspin         =2</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> /</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> &electrons</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    conv_thr      = 1e-8</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    mixing_beta   = 0.7</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">    mixing_mode   ='plain'</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> /</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">ATOMIC_SPECIES</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">  Fe   55.847    Fe.pw91-spn-rrkjus.UPF</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">ATOMIC_POSITIONS alat</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">  Fe    0.0  0.0  0.0</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">K_POINTS {automatic}</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">  8 8 8  1 1 1</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> </span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">Dr Virginie Trinite</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">---------------------------------------</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">Modeling Infra-Red Lasers and Detectors</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US">III-V Lab THALES Research&  Technology, France</span></p><p><span style="font-size:10.5pt;font-family:Consolas;" lang="EN-US"> </span></p><p><span style="font-size:10.5pt;font-family:Consolas;">Campus Polytechnique</span></p><p><span style="font-size:10.5pt;font-family:Consolas;">1, avenue Augustin Fresnel 91767 Palaiseau cedex France</span></p><p><span style="font-size:10.5pt;font-family:Consolas;">Phone : +33 (0)1.69.41.58.20</span></p><p> </p></div></div><p> </p></div><p>,</p><p>_______________________________________________<br /> Pw_forum mailing list<br /> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br /> <a href="http://pwscf.org/mailman/listinfo/pw_forum">http://pwscf.org/mailman/listinfo/pw_forum</a></p></blockquote></div>,<p>_______________________________________________<br />Pw_forum mailing list<br /><a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br /><a href="http://pwscf.org/mailman/listinfo/pw_forum">http://pwscf.org/mailman/listinfo/pw_forum</a></p></blockquote>