[Wien] [SPAM?] Re: same k vector, different eigenvectors

Wed Apr 26 12:04:43 CEST 2023

The version of fort that you are using is an antique! I certainly suggest
updating to (free) OneAPI. I use 2021.1.1 since the later versions seem
slightly slower. Also, what Wien2k?

In terms of compiler options, I suggest using/adding
-O2 -m64 -msse4.2 -ip -prec-div -r8 -pc80 -fpconstant -prec-sqrt
-fimf-precision=high -no-complex-limited-range -no-fast-transcendentals
-no-ftz -sox -fma

You will have to check if your antique accepts all of these.

Also, how was your Wien2k compiled (details), and are you using elpa? If
you are, how was elpa compiled. What mkl?

N.B., you should check case.output1?? for the relevant vectors to ensure
that there is not something subtly different in the linearization energy.
Then the same for lapwso.

N.N.B., I have not checked what your numbers mean, so I cannot comment on
the accuracy. At best it is 1 part in 10**15, sometimes only 1 part in
10**13.

--
Professor Laurence Marks (Laurie)
Department of Materials Science and Engineering, Northwestern University
www.numis.northwestern.edu
"Research is to see what everybody else has seen, and to think what nobody
else has thought" Albert Szent-Györgyi

On Wed, Apr 26, 2023, 04:11 曹迎迎 <yycao at iphy.ac.cn> wrote:

> Dear Prof. Blaha,
>
>   Thank you for your prompt reply.
>
>   I currently use ifort + mkl (composer_se_2013.0.079).
>
>   The difference between the two momentum matrices is in the same order
> of magnitude as the momentum matrix itself, so it does not seem like
> noise. The momentum matrix for 2-kpt case is as follows:
>
> ...
>
>    KP:     1 NEMIN NEMAX :     1   28 dE: -10.0   3.0 K:       101
>
>  1  1  2.250347E-16  5.044130E-19  1.784245E-16 -5.197770E-19
> 1.112087E-15 -3.976961E-18    0.00000000
>  1  2 -7.138923E-17 -1.592665E-17 -2.366714E-17  3.086282E-17
> -2.926389E-17  4.154997E-17    0.19969467
>  1  3  3.840733E-01  1.521339E-01  1.521339E-01 -3.840733E-01
> 6.435560E-13 -6.350574E-13    2.43558099
>  1  4 -3.607201E-13 -3.502108E-13 -8.289942E-13 -8.124813E-13
> 4.113744E-01  3.991154E-01    2.46843878
>  1  5  2.049534E-01 -3.719653E-01  3.719653E-01  2.049534E-01
> 9.441647E-13  1.008950E-13    2.52112809
>  1  6 -1.015254E-13 -9.906789E-14 -2.730212E-13 -2.625597E-13
> 1.258610E-01  1.221065E-01    2.65672779
> ...,
>
> while the one for 101-kpt case is
>
> ...
>
>    KP:     4 NEMIN NEMAX :     1   28 dE: -10.0   3.0 K:       101
>  1  1  3.804507E-16  1.545496E-19  3.073192E-16 -1.485961E-19
> -1.441558E-15 -8.622593E-20    0.00000000
>  1  2 -9.188564E-17  1.787862E-17 -1.536060E-16 -2.111259E-17
> 7.786885E-18  3.037665E-17    0.19969467
>  1  3  6.091761E-04  4.131061E-01  4.131061E-01 -6.091761E-04
> 1.374420E-12  1.209410E-12    2.43558099
>  1  4 -1.420821E-12  8.882743E-13 -1.590622E-12  9.956160E-13
> 4.859826E-01 -3.038798E-01    2.46843878
>  1  5 -3.935275E-01  1.596876E-01 -1.596876E-01 -3.935275E-01
> -1.655225E-12 -7.933061E-13    2.52112809
>  1  6 -4.270845E-13  2.667982E-13 -4.438459E-13  2.776656E-13
> 1.486812E-01 -9.297802E-02    2.65672779
> ....
>
>  However, it appears that the symmetrized squared momentum matrix
> elements do have only numerical noise. They are:
>
> ...
>
>    KP:     1 NEMIN NEMAX :     1   28 dE: -10.0   3.0 K:       101
>
>       1   1 0.412382E-31 0.412382E-31 0.123675E-29 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   2 0.343136E-32 0.343136E-32 0.258278E-32
> 0.000000E+00-0.106911E-49 0.000000E+00
>       1   3 0.170657E+00 0.170657E+00 0.817462E-24 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   4 0.800062E-24 0.800062E-24 0.328522E+00
> 0.000000E+00-0.315544E-29 0.000000E+00
>       1   5 0.180364E+00 0.180364E+00 0.901627E-24 0.000000E+00
> 0.157772E-29 0.000000E+00
>       1   6 0.818000E-25 0.818000E-25 0.307510E-01 0.000000E+00
> 0.000000E+00 0.000000E+00
> ...
>
> and
>
> ...
>
>    KP:     4 NEMIN NEMAX :     1   28 dE: -10.0   3.0 K:       101
>
>       1   1 0.119594E-30 0.119594E-30 0.207809E-29 0.000000E+00
> 0.547382E-47 0.000000E+00
>       1   2 0.164016E-31 0.164016E-31 0.983377E-33 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   3 0.170657E+00 0.170657E+00 0.335170E-23 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   4 0.316455E-23 0.316455E-23 0.328522E+00 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   5 0.180364E+00 0.180364E+00 0.336911E-23 0.000000E+00
> 0.000000E+00 0.000000E+00
>       1   6 0.263840E-24 0.263840E-24 0.307510E-01 0.000000E+00
> 0.000000E+00 0.000000E+00
> ....
>
> I am wondering if this means that i can't calculate the transverse
> conductivity and Berry curvature from the momentum matrix output by OPTIC?
>
>   Thank you for your response and support. I look forward to hearing back
> from you soon.
>
>
> Best regards,
>
> Yingying Cao
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20230426/fe61b66a/attachment.htm>