Page 1 of 1

Regarding GW+BSE calculations using yambo-5.0.4

Posted: Tue Aug 09, 2022 10:58 am
by Jaspreet Singh
Dear Developers,
I am quiet new to the yambo calculations. As told in the previous post, I have updated my yambo code. Now, I want to obtain the GW band structure and optical absorbance spectra using GW+BSE method. I have done the GW calculations and obtain the band structure. However, the results obtained are not quite good because there is a no change in GW band gap from HSE band gap. Also, the BSE calculations are completed very fast (within minutes). Can, anybody provide the input files for GW and BSE calculations that contain the required parameters (including the parallelization info) to obtain the accurate results. I will than update the parameters according to my system.
Regards

Re: Regarding GW+BSE calculations using yambo-5.0.4

Posted: Thu Aug 11, 2022 7:35 am
by claudio
Dear Jaspreet

please have a look to the wiki webpage, where there is a description of all parameters one has to converge in GW and BSE:

https://www.yambo-code.eu/wiki/index.ph ... n_parallel

the same for the BSE

https://www.yambo-code.eu/wiki/index.ph ... parameters

and if it can be useful look at my blog where I suggest decent parameters for the calculations:

Reasonable parameters for Yambo calculations:
http://www.attaccalite.com/reasonable-p ... culations/

let me know if you have other questions

best
Claudio

Re: Regarding GW+BSE calculations using yambo-5.0.4

Posted: Wed Aug 17, 2022 5:51 am
by Jaspreet Singh
Dear Claudio,
I have done the GW calculations from which i have obtained the band structure (gets completed in few days on 16 cores), however the BSE calculations are completed very fast. Is it correct or I'm doing some mistake. Note that I'm using the yambo version 5.0.4.
Thanks

Re: Regarding GW+BSE calculations using yambo-5.0.4

Posted: Wed Aug 17, 2022 10:01 am
by Daniele Varsano
Dear Jaspreet,

the computational cost of GW and BSE it depends on the used parameters and you need to check that your calculation is converged.
If you post your input/report/log files we can have a look.

Best,
Daniele