Dear Nick, How did you build the OpenMP version of QE? I was unable to find any reference to such a flavor for QE nor to the omp header in source files.
Regards, Reinaldo Reinaldo Pis Diez PhD in Chemistry Principal Researcher - CONICET (Argentina) Associated Professor - UNLP (Argentina) On 01/27/2016 03:33 PM, Nick Wilson wrote: > >> On 26 Jan 2016, at 20:02, Paolo Giannozzi <[email protected] >> <mailto:[email protected]>> wrote: >> >> >> One should obtain the same numbers - within the errors due to >> roundoff, though - in serial, OpenMP, MPI execution, and on >> different machines, and with different compilers and >> mathematical libraries. >> >> >> ... actually, only at exact convergence, at least when using >> default wavefunction initialization: small differences come from >> different randomization of atomic wavefunctions. These however >> disappear as convergence is made tighter (but sometimes they don’t > > > Thanks Paolo & Stefano, that’s very useful information. > > The convergence threshold was at 1.0e-9 and when I tightened it to > 1.0e-12 it improved the agreement significantly. > > Best regards, > Nick > > > _______________________________________________ > Pw_forum mailing list > [email protected] > http://pwscf.org/mailman/listinfo/pw_forum _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
