You sent an output from "projwfc.x", clearly showing that parallel diagonalization (ScaLAPACK) is used for overlap matrices, whose size is natomwfc x natomwfc. Iterative parallelization used in self-consistency is a different story.
Paolo On Thu, Feb 14, 2019 at 9:39 AM LEUNG Clarence <[email protected]> wrote: > Dear Paolo, > > Thanks for your reply. > But I am not sure, You mean that diagonalization in scf uses 'cg' ? In my > scf, diagonalization used the default option 'david' . > > Clarence > ------------------------------ > *发件人:* users <[email protected]> 代表 Paolo > Giannozzi <[email protected]> > *发送时间:* 2019年2月14日 16:28 > *收件人:* Quantum Espresso users Forum > *主题:* Re: [QE-users] PDOS for isolated gas molecule > > > On Thu, Feb 14, 2019 at 9:18 AM LEUNG Clarence <[email protected]> > wrote: > > When I calculate the PDOS for isolated gas molecule by projwfc.x , it stop > with an error : > > > > Error in routine blk2cyc_zredist (1): > nb less than the number of proc > > > you have 2x2 matrices (natomwfc x natomwfc), don't use parallel > diagonalization. > Paolo > -- > Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche, > Univ. Udine, via delle Scienze 208, 33100 Udine, Italy > Phone +39-0432-558216, fax +39-0432-558222 > > _______________________________________________ > users mailing list > [email protected] > https://lists.quantum-espresso.org/mailman/listinfo/users -- Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche, Univ. Udine, via delle Scienze 208, 33100 Udine, Italy Phone +39-0432-558216, fax +39-0432-558222
_______________________________________________ users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
