In data 21 ottobre 2009 alle ore 10:41:33, Jiri Houska <jhouska at kfy.zcu.cz> ha scritto:
> Dear Lorenzo, > thanks for the answer. The relaxation is from scratch (not restarted), > QE 4.0.3, running on cluster consisting of quadcore Xeon E5472 (3 GHz, > 12 MB cache), happened for both (1) fixed occupations [used originally > for material with a band gap] and (2) cold smearing. Dear Jiri, what happened in both cases? The i/o error or the algebra error? As far as I know, the Berry-phase method does not work with pools (for implementation reasons), also I'm not quite sure it works with smearing, and with metals in general (I think the theory only applies to insulators). The davcio error you have experienced may happen if you restart a calculation with a different number of pools. On the other hand, the zgefa error may be caused by dipfield+smearing or dipfield+pools. You could check if one of the aforementioned conditions is verified in your setup. cheers -- Lorenzo Paulatto SISSA & DEMOCRITOS (Trieste) phone: +39 040 3787 511 skype: paulatz www: http://people.sissa.it/~paulatto/ *** save italian brains *** http://saveitalianbrains.wordpress.com/
