k-point symmetry/reduction

[Léon Luntadila Lufungula]

Dear all,

I have been testing some settings in my CP2K input file regarding symmetry. I got most of them working but I am getting a CPASSERT error when trying to use the SYMMETRY keyword for the k-points. I just stumbled upon the FAQ reqarding k-points and it states that "Only FULL_GRIDs of k-points may be used, no reduction due to symmetry is supported.". So I'm wondering why FULL_GRID and SYMMETRY are even an option if we can't use them? Also, will these be implemented at some point or is this not on the priorities list?

Kind regards,

Léon

[Frederick Stein]

Dear Leon,

Please excuse the missing implementation. I agree that it is unfortunate to have keywords without functionality available. I guess these keywords were added with the intent to implement the functionality later which actually never happened.

To our understanding, an extension of the k-point-functionality in CP2K to be competitive with other codes requires a larger amount of time (probably years) and was traditionally not the main interests of the CP2K developers (CP2K is mostly used for simulations where k-points will not be of help anyways).

In our group, we do have an eye on it but right now not the capacity to work on it and I am not aware of other groups working on general k-point-functionalities. I am aware of a development regarding TDDFT with k-points in CP2K although the respective code has never been added to CP2K's main repository.

Kind regards,

Frederick