Dear all,
I am running electron-phonon coupling as described in Si example. I find if I take the random q ponit a larger number in 02 ph and 04 elph process, the phonon calculation will slowdown because the huge q point. Do we have an idea to accelerate the phonon calculation for this step since we need to test
the number of the random q point we used.
Thanks!
Shudong
Speed in the electron-phonon calculation
Moderators: palful, amolina, mbonacci
-
- Posts: 299
- Joined: Fri Apr 09, 2010 12:30 pm
- amolina
- Posts: 135
- Joined: Fri Jul 15, 2011 11:23 am
- Location: Valencia, Spain
- Contact:
Re: Speed in the electron-phonon calculation
Dear Shudong,
One way to accelerate the electron-phonon calculation is to parallelize in q-points. That is, running an individual job for each q-point (using a bash or python script).
Best regards,
Alejandro.
One way to accelerate the electron-phonon calculation is to parallelize in q-points. That is, running an individual job for each q-point (using a bash or python script).
Best regards,
Alejandro.
Alejandro Molina-Sánchez
Institute of Materials Science (ICMUV)
University of Valencia, Spain
Institute of Materials Science (ICMUV)
University of Valencia, Spain
-
- Posts: 299
- Joined: Fri Apr 09, 2010 12:30 pm
Re: Speed in the electron-phonon calculation
Dear Alejandro,
I am trying to use image to accelerate the phonon calculation but it faild when I set "fildvscf = 'dvscf'".
If I omit this item, everything is OK. My question is does the fildvscf must be set in the phonon calculation for next Yambo procedure?
Thanks!
Shudong
I am trying to use image to accelerate the phonon calculation but it faild when I set "fildvscf = 'dvscf'".
If I omit this item, everything is OK. My question is does the fildvscf must be set in the phonon calculation for next Yambo procedure?
Thanks!
Shudong