[QE-users] QE 6.5 versus QE 7.2

[Giovanni Cantele]

Dear all,

I'm redoing some old calculations that were carried out using QE 6.5,
few years ago.
The new calculations use the latest QE version.

Input files are exactly the same.

By analysing the differences of the output files, for example for an scf
calculation, I noticed what follows:

< !    total energy              =    -390.19169099 Ry
<      estimated scf accuracy    <       0.00000008 Ry
<      smearing contrib. (-TS)   =       0.00000000 Ry
<      internal energy E=F+TS    =    -390.19169099 Ry
---
> !    total energy              =    -390.18965771 Ry
>      estimated scf accuracy    <       0.00000042 Ry
>      smearing contrib. (-TS)   =       0.00273433 Ry
>      internal energy E=F+TS    =    -390.19239205 Ry

and following lines concerning contributions to total energy
<      one-electron contribution =    -119.48113307 Ry
<      hartree contribution      =     120.72351434 Ry
<      xc contribution           =     -99.27362121 Ry
---
>      one-electron contribution =    -119.47953593 Ry
>      hartree contribution      =     120.72030987 Ry
>      xc contribution           =     -99.27271495 Ry


While I would not expect to exactly the same numbers in the two cases, it
seems to me that the total energy differences are too large to be justified
with numerical oscillations, parallelization over a different number of
cores and so on:
 -390.19169099 - (-390.18965771) ~ -2 mRy ~ 28 meV

Total forces
      Total force =     0.003905     Total SCF correction =     0.000228
---
>      Total force =     0.002307     Total SCF correction =     0.000701



Stress tensor
<           total   stress  (Ry/bohr**3)                   (kbar)     P=
    -2.14
<   -0.00001057   0.00000000  -0.00000215           -1.55        0.00
-0.32
<    0.00000000  -0.00001982   0.00000000            0.00       -2.92
 0.00
<   -0.00000215   0.00000000  -0.00001321           -0.32        0.00
-1.94
---
>           total   stress  (Ry/bohr**3)                   (kbar)     P=
    -2.30
>   -0.00001231   0.00000000  -0.00000064           -1.81        0.00
-0.09
>    0.00000000  -0.00001774  -0.00000000            0.00       -2.61
-0.00
>   -0.00000064  -0.00000000  -0.00001687           -0.09       -0.00
-2.48



Fermi energy (in a subsequent nscf calculation)
<      the Fermi energy is     7.3474 ev
<      (compare with:     7.3589 eV, computed in scf)
---
>      the Fermi energy is     6.4844 ev
>      (compare with:     6.4868 eV, computed in scf)
(for Fermi energy I'm a little less worried, since both values fall within
the gap of the material).


Do you have any guesses about what could be the reason for these different
results? Do you think that, as it seems to me, these differences are not
that small?

I thank you in advance.
Giovanni

-- 

Giovanni Cantele, PhD
CNR-SPIN
c/o Dipartimento di Fisica
Universita' di Napoli "Federico II"
Complesso Universitario M. S. Angelo - Ed. 6
Via Cintia, I-80126, Napoli, Italy
e-mail: giovanni.cantele at spin.cnr.it <giovanni.cantele at spin.cnr.it>
Phone: +39 081 676910
Skype contact: giocan74

ResearcherID: http://www.researcherid.com/rid/A-1951-2009
Web page: https://sites.google.com/view/giovanni-cantele/home
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230519/cf19c90d/attachment.html>