My Community

VASP team,

I have a quick question, I have performed a few GW calculations Be, and I am wondering why the self-energy values for the low-scaling (G0W0R) method are all negative, while the self-energies for the traditional GW (EVGW0) are positive below the Fermi energy and negative above.

Thanks for your help!

Please post both calculations according to the forum guidelines.

Sorry about that, I have attached the calculations. The EVGW0.OUTCAR file was modified due to its size.

In both calculations you are using very different parameters.
For the low-scaling GW you are using a finite-temperature formalism (LFINITE_TEMPERATURE=.TRUE.), while for the EVGW0 calculation not. This is correct, since this formalism is only available for the low-scaling method. You also have a metal and for that you have to use a finite-temperature fomalism to obtain correct results. Additionally the number of frequency gridpoints (NOMEGA) is also different in both.

So if you want to compare the two methods you should try it on an insulator, otherwise for metals please use the results from the low-scaling GW method including the finite-temperature formalism.