GW₀

[kxzhou]

Dear all,
When performing GW₀ self-consistent iterations, I have a couple of questions:
1. Can all the files generated from the initial G₀W₀ calculation (such as ndb.HF_and_locXC, ndb.dipoles, ndb.pp*, ndb.cutoff, ndb.RIM, ndb.RIM_W) be directly reused for each subsequent iteration, with the exception of ndb.QP? To confirm my understanding: in each iteration, is the only component being actively recalculated the correlation self-energy Σc (by updating the Green's function G with the latest quasiparticle energies), which is then used to generate a new ndb.QP file?
2. Additionally, regarding the quasiparticle bands: how many bands need to be included in the calculation for each iteration? Is it possible to only calculate a few bands around the Fermi level? I've noticed this step is extremely memory-intensive.

[Daniele Varsano]

Dear Kexin,

please sign your post with your name and affiliation. This is a rule of the forum and you can do once for all by filling the signature in your profile.

1) Yes, if the screening is kept fixed, only the ndb.QP is updated.

2) Yes, you can include few bands across the gap (QPkrange), but pay attention as reported in the tutorial:

You should perform some tests increasing the bands range and see if you converge to the same result.
Notice that at each iteration the GW correction is fitted and applied also to the bands not included in the bands range. Do not use non-continuous bands/k-points range this could give wired results and the code does not check against these errors.

Best,
Daniele