GW0 convergence vs NOMEGA

[d.pshenay]

Dear all,

I have two questions concerning convergence vs NOMEGA in GW0 calculations:

1) I'd like to reach convergence vrt NOMEGA. E.g. I set NBANDS to maximum allowed for given Ecut. Then, if I understand right, if I will increase NOMEGA at fixed k-sampling, than from certain value I shoud see not the convergence but noise increase. So the only straitforward way is to increase k-sampling and repeat NOMEGA variation and compare results. Am I right?

Taking into account GW0 computing time, it should be quite long calculations. So, another option is to use recommended values. But there is another question:

2) It is recommended to set NOMEGA<=100 for 4x4x4 k-points/atom and <=200-300 for 6x6x6 points/atoms. I expected that increasing number of atoms means increasing of unit cell, then I can decrease number of points in k-sampling. But the number of atoms in recepy is in denominator (k-points/atom). I am not sure, why? E.g. if I have 1 atom per cell then I can use NOMEGA=100 for 4x4x4 grid, what will be recommended NOMEGA for 2 atoms per cell and same 4x4x4 grid?

Best regards,
Dmitry

[support_vasp]

Hi,

We're sorry that we didn’t answer your question. This does not live up to the quality of support that we aim to provide. The team has since expanded. If we can still help with your problem, please ask again in a new post, linking to this one, and we will answer as quickly as possible.

Best wishes,

VASP