Page 2 of 2

Re: evGW

Posted: Fri Feb 27, 2026 9:23 am
by Daniele Varsano
Dear Xiangjun,
are ndb.pp*` files themselves affected by `rim_cut` and `rim_w`
Yes, they are!
I also have another question. In my G0W0 calculation, I used `rim_cut` and `rim_w`. Do I still need to use them in G1W0?
I presume you are dealing with a 2D system.
As you are not updating W the previously calculated ndb.pp and ndb.rimw are read and not recalculated. Anyway the flags should be kept for compatibility, also as you are recalculating the HF part rim_cut should be kept to truncate the coulomb potential.

If you are finding differences omitting the rim_cut and rim_W it is probably because you are recalculating the HF part without coulomb cutoff. You can inspet your report file and check what has been read and what has been recalculated by looking at the [RD] and [WR] when the header of the databases are reported.

Best,

Daniele