Page 1 of 1

Problem in compiling VASP 5.2

Posted: Mon Feb 14, 2011 4:16 am
by Amritendu
Hi all:
we complied vasp 5.2 using mpif90 compiler. The compliation was done successfully aparently and we submitted job. In the OSZICAR file we got the following:
RMM: 20 -0.287820647200E+03 0.10207E-03 -0.18326E-02 56590 0.574E-01 0.206E-01
RMM: 21 -0.287820473355E+03 0.17384E-03 -0.42106E-03 74646 0.181E-01 0.902E-02
RMM: 22 -0.287820422159E+03 0.51196E-04 -0.46393E-04 69666 0.528E-02 0.697E-02
RMM: 23 -0.287820337687E+03 0.84472E-04 -0.70124E-04 63375 0.828E-02 0.197E-02
RMM: 24 -0.287820337266E+03 0.42105E-06 -0.66046E-05 66425 0.323E-02 0.107E-02
RMM: 25 -0.287820337321E+03 -0.55177E-07 -0.19049E-05 70314 0.109E-02 0.555E-03
RMM: 26 -0.287820336974E+03 0.34726E-06 -0.30842E-06 74010 0.431E-03 0.257E-03
RMM: 27 -0.287820336966E+03 0.81964E-08 -0.14404E-06 70878 0.304E-03 0.132E-03
RMM: 28 -0.287820336927E+03 0.39421E-07 -0.20480E-07 68361 0.148E-03 0.708E-04
RMM: 29 -0.287820336920E+03 0.69940E-08 -0.85846E-08 72918 0.647E-04 0.184E-04
RMM: 30 -0.287820336920E+03 -0.75124E-09 -0.12587E-08 67242 0.322E-04 0.177E-04
RMM: 31 -0.287820336920E+03 0.52023E-09 -0.25995E-09 68478 0.138E-04 0.353E-05
RMM: 32 -0.287820336920E+03 -0.83674E-10 -0.73056E-10 68103 0.603E-05 0.291E-05
RMM: 33 -0.287186498798E+03 0.63384E+00 0.75268E+00 61690 0.300E-05 0.871E-02

it was surprising to find the dE value increases 10 order after iteration 32. Could anyone please help us sort out the problem

Problem in compiling VASP 5.2

Posted: Fri Mar 11, 2011 6:59 pm
by chelman
I suggest you to explain a little bit more you system. Is that happen every runs or only in in this system?. If that the case what kind of system do you have? metal, slab, etc... Then the community can give you some suggestions.
bye

Problem in compiling VASP 5.2

Posted: Mon Mar 14, 2011 10:30 am
by alex
You are spanning a range of at least 13 orders of magnitude in energy. Why do you use such tight criteria?

The minimizer doesn't find the minima within your limits, makes a big jump and searches somewhere else.

cheers,

alex