[Wien] orb calculations on IBM-AIX and Linux give different results

[Stefaan Cottenier]

Are you sure you started both calculations in exactly the same  
circumstances. No dmat or orb files from previous runs or from a  
restore_lapw lingering around? LDA+U should be deterministic, but the  
final result is determined by the starting clmsum*, dmat* and orb*.

Stefaan

Quoting Hong Jiang <jiang at fhi-berlin.mpg.de>:

> Dear Wien2k developers and users,
> Recently I have found something really strange about orb (LDA+U, nsic=1)
> calculations.
> When I did some test calculations using MnO in ferromagnetic phase (
> using runsp_lapw), I found results from Linux and AIX are significantly
> different.
> On Linux, U_d=0.3 Ry. (J_d is always taken as zero) is already enough to
> give a band gap in the majority spin channel. But on AIX 5.3, even using
> U_d=0.6 Ry. still does not give a gap. The total energy are also different.
> On the other hand, a pure runsp_lapw without -orb does give identical
> results on Linux and AIX.
>
> Do you have any idea about that?
>
> Thank you very much in advance!
>
> Hong
>
>
>
>
> PS: more details
>
> nsic=1, U_d=0.6 Ry, J_d=0.0
>
> +++++++++++++++++++++++++++++++ Linux +++++++++++++++++++++++++++++
>
> TOTAL ENERGY IN Ry = -2462.537254
>
> Insulator, EF-inconsistency corrected
> Bandranges (emin - emax):
> :
> :BAND : 5 -0.003969 0.293250
> :BAND : 6 0.108061 0.293250
> :BAND : 7 0.172152 0.293250
> :BAND : 8 0.572991 0.844173
> :BAND : 9 0.736397 0.942032
> :BAND : 10 0.809698 0.958555
>
> :FER : F E R M I - ENERGY(TETRAH.M.)= 0.43250
>
>
> ++++++++++++++++++++++++++++++++++++++++++ AIX ++++++++++++++++++++
> TOTAL ENERGY IN Ry = -2461.520991
>
> Bandranges (emin - emax):
> :BAND : 5 0.001049 0.303735
> :BAND : 6 0.115155 0.303735
> :BAND : 7 0.175177 0.303735
> :BAND : 8 0.574089 0.803644
> :BAND : 9 0.692973 0.828036
> :BAND : 10 0.722234 0.841208
>
> :FER : F E R M I - ENERGY(TETRAH.M.)= 0.60770
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>



-- 
Stefaan Cottenier
Instituut voor Kern- en Stralingsfysica
K.U.Leuven
Celestijnenlaan 200 D
B-3001 Leuven (Belgium)

tel: + 32 16 32 71 45
fax: + 32 16 32 79 85
e-mail: stefaan.cottenier at fys.kuleuven.be


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm