Hello, I am currently trying to calculate the frequency-dependent dielectric function of a cluster using GW+BSE calculations. When performing GW+BSE calculations in VASP, is it possible to compute clusters using only a single Gamma k-point? I am encountering an issue where the calculation gets killed (error: signal 9) during the GW step. After investigating the cause, it seems to be related to a memory shortage, so I have tried various solutions (increasing nodes, parallelizing the calculation, etc.), but I've not been able to resolve the issue yet. I am starting to wonder if this problem occurs because VASP's GW+BSE calculations may not be fundamentally suited for cluster calculations.

Additionally, I am wondering if it is possible to compute dielectric function values in the high-energy range (~100 eV) using GW+BSE calculations.

I would appreciate your insight on this. Thank you in advance.