Page 1 of 1

Questions about Real-time BSE calculations?

Posted: Thu Jan 16, 2020 7:17 am
by Dean
Dear all,
I am doing the Real-time BSE calculations of 2D InSe following the TUTORIAL http://www.attaccalite.com/lumen/real_time_bse.html.
However, when I try to perform calculations of the collisions, the Yambo stopped without any error messages, and the output file r-2D_em1s_collisions ends with:

"[04] Static Dielectric Matrix
=============================

[RD./SAVE//ns.kb_pp_pwscf]----------------------------------
Fragmentation :yes
- S/N 008119 -------------------------- v.04.03.00 r.00129 -

[WARNING] [x,Vnl] slows the Dipoles computation. To neglect it rename the ns.kb_pp file "

The input and output files have been attached.
I would be very glad for any response.
With regards.

Re: Questions about Real-time BSE calculations?

Posted: Thu Jan 16, 2020 8:53 am
by claudio
Dear Dean

please try this new version of the tutorial with yambo 4.5 and let us know if you have the same problems

http://www.yambo-code.org/wiki/index.ph ... ion_(TDSE)

best
Claudio

Re: Questions about Real-time BSE calculations?

Posted: Fri Jan 17, 2020 3:25 am
by Dean
Thanks very much.
I will try it.

Re: Questions about Real-time BSE calculations?

Posted: Fri Jan 17, 2020 5:35 am
by Dean
Dear Claudio,
I do some tests and find that:
1. The number of mpi threads results in the CRASH. For my case, 23 mpi threads works well, but 24 results in the CRASH.
2. When I set ExtF_kind= "DELTA", output file shows the "[WARNING] Non-linear optics properties can be obtained with SOFTSIN/SIN fields only!"
Does it matter?
3. For 2D materials, the use of Random Integration Method and a truncated Coulomb potential is still available in real-time BSE or nonlinear calculations?

Thanks in advance.

Re: Questions about Real-time BSE calculations?

Posted: Fri Jan 17, 2020 9:06 am
by claudio
Dear Dean

> 1. The number of mpi threads results in the CRASH. For my case, 23 mpi threads works well, but 24 results in the CRASH.

I will check this issue, in practise when you run the tutorial example on 24 core it crashes?
something like

mpirun -np 24 yambo_nl -F input.in

2. When I set ExtF_kind= "DELTA", output file shows the "[WARNING] Non-linear optics properties can be obtained with SOFTSIN/SIN fields only!"
Does it matter?

no, does not matter

3. For 2D materials, the use of Random Integration Method and a truncated Coulomb potential is still available in real-time BSE or nonlinear calculations?

yes

best
Claudio

Re: Questions about Real-time BSE calculations?

Posted: Sun Jan 19, 2020 1:22 pm
by Dean
Dear Dr. Claudio,
Thanks for your reply.
When I run my own job, it crashes.
In detail, when I use the commond "mpirun -np 24 yambo_nl -F input.in", it crashes. When I use the commond "mpirun -np 23 yambo_nl -F input.in", it does not crashes.

Re: Questions about Real-time BSE calculations?

Posted: Tue Jan 21, 2020 5:43 pm
by Davide Sangalli
Dear Dean,
could you please attach the report file (r-*) of the run and the file "report" which is inside the config folder in the source of the yambo code ?
Thanks.

Kind regards,
Davide

Re: Questions about Real-time BSE calculations?

Posted: Thu Jan 23, 2020 12:42 pm
by Dean
Dear Dr. Davide,
Thanks for your reply.
The files you mentioned are attached. Please check them.