Calculation stopping with "BAD TERMINATION" -- RIM-W

Concerns issues with computing quasiparticle corrections to the DFT eigenvalues - i.e., the self-energy within the GW approximation (-g n), or considering the Hartree-Fock exchange only (-x)

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano

Post Reply
User avatar
matteo_dalessio
Posts: 5
Joined: Wed Jun 22, 2022 6:23 pm

Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by matteo_dalessio » Sun Jun 26, 2022 8:23 pm

Dear all,

I'm calculating quasiparticle corrections to DFT band structure for a 2D semiconductor and I'm using Yambo 5.1.
At the moment I'm studying convergence of the gap at Γ with respect to BndsRnXp and NGsBlkXp, but before that I'm trying to do a test calculation with the input gw_ppa.in here attached. The problem I run into is that the calculation stops at some point during the Dyson equation (I'm using Newton solver) with no errors in the report or in the log. I'm using slurm to run the calculation and the only errors I find are in the job.out file. This is how I run the job

Code: Select all

#! /bin/bash

#SBATCH --ntasks-per-node=16
#SBATCH --cpus-per-task=1
#SBATCH -o gw_ppa_jobcm.out
#SBATCH --job-name='gw_ppa_test'

mpirun -np $SLURM_NTASKS ~/Software/yambo-5.1.0/bin/yambo -F gw_ppa.in -J gw_ppa_test
and the errors in the job.out are

=================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= RANK 15 PID 1411080 RUNNING AT cm01
= KILLED BY SIGNAL: 9 (Killed)
=================================================================================


Any help for this issue would be really appreciated. Thank you,

Matteo
You do not have the required permissions to view the files attached to this post.
Last edited by matteo_dalessio on Tue Jun 28, 2022 11:08 am, edited 1 time in total.
Matteo D'Alessio
Università degli Studi di Modena e Reggio Emilia
Dipartimento di Scienze Fisiche, Informatiche e Matematiche

User avatar
matteo_dalessio
Posts: 5
Joined: Wed Jun 22, 2022 6:23 pm

Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by matteo_dalessio » Tue Jun 28, 2022 11:08 am

Dear developers,

I ran several tests changing parallelisation parameters and even on other machines because I thought it was a memory issue, but nothing changed. Eventually I removed the RIM_W method keeping the rest of the input unchanged and the calculation runs without any problem.
The relevant files are attached.

Thank you for your attention,

Matteo
You do not have the required permissions to view the files attached to this post.
Matteo D'Alessio
Università degli Studi di Modena e Reggio Emilia
Dipartimento di Scienze Fisiche, Informatiche e Matematiche

sdwang
Posts: 287
Joined: Fri Apr 09, 2010 12:30 pm

Re: Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by sdwang » Wed Jun 29, 2022 10:59 am

Dear all,
I met the same problem and I think it is not the memory problem at least.

shudong
S. D. Wang
IMU,HOHHOT,CHINA
E-mail: sdwang@imu.edu.cn

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by Daniele Varsano » Mon Jul 04, 2022 9:24 am

Dear Matteo and Shudong,
thanks for reporting, we will try to reproduce and spot the problem.
Best,
Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

Alberto Guandalini
Posts: 8
Joined: Thu Jun 24, 2021 9:10 am

Re: Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by Alberto Guandalini » Mon Jul 04, 2022 3:47 pm

Dear Matteo and Shudong,
thank you for your reports.

We find a bug in the RIM-W implementation in the 5.1 version of the code.
The bug has been fixed and will be available in the next patch of the code.

You have two possibilities: waiting for the new patch to be released, or to solve the bug yourself in the
version 5.0. It is very easy and explained in the following

-------------------
In the file src/qp/QP_interpolate_W.F, move the line 283:

YAMBO_FREE(f_coeff)

between the line 360 and 361 (just before the end of the subroutine).
-------------------

Best,
Alberto

User avatar
matteo_dalessio
Posts: 5
Joined: Wed Jun 22, 2022 6:23 pm

Re: Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by matteo_dalessio » Mon Jul 04, 2022 5:22 pm

Thank you Alberto, I'm glad it has been solved :)

Best,

Matteo
Matteo D'Alessio
Università degli Studi di Modena e Reggio Emilia
Dipartimento di Scienze Fisiche, Informatiche e Matematiche

sdwang
Posts: 287
Joined: Fri Apr 09, 2010 12:30 pm

Re: Calculation stopping with "BAD TERMINATION" -- RIM-W

Post by sdwang » Tue Jul 05, 2022 2:51 am

Dear Daniele and Alberto,
I have fixed the problem as suggested by Alberto, and now it works!
Thanks for your kindly help!

Best,


Shudong
S. D. Wang
IMU,HOHHOT,CHINA
E-mail: sdwang@imu.edu.cn

Post Reply