Page 1 of 1

Incomplete simulation of single point SCF calculation

Posted: Sat May 07, 2022 2:14 pm
by aniket_singha
Slurm job - exit status zero
Job is a Single point SCF calculation
Does not show any error but does not compute and exits

I am attaching the INCAR, KPOINT, POSCAR, POTCAR, OUTCAR, slurm script submit,sh
PFA


Last lines
k-point 1 : 0.0000 0.0000 0.0000 plane waves: 990747

maximum and minimum number of plane-waves per node : 990747 990747

maximum number of plane-waves: 990747
maximum index in each direction:
IXMAX= 97 IYMAX= 51 IZMAX= 47
IXMIN= -97 IYMIN= -51 IZMIN= -47


serial 3D FFT for wavefunctions
parallel 3D FFT for charge:
minimum data exchange during FFTs selected (reduces bandwidth)

minimum data exchange during FFTs selected (reduces bandwidth)

Posted: Sun May 08, 2022 8:23 am
by aniket_singha
I have attached the zip file where INCAR, POSCAR, POTCAR, K-POINTS and slurm batchscript is present. It is working well for simple Si structure.
but this single point calculation stops with slurm output zero meaning successful implementation.
The system contains around 440 atoms. I am looking for finding out the plane wave cutoff and k-point sampling so as to go for structural optimization.

k-point 1 : 0.0000 0.0000 0.0000 plane waves: 990747

maximum and minimum number of plane-waves per node : 990747 990747

maximum number of plane-waves: 990747
maximum index in each direction:
IXMAX= 97 IYMAX= 51 IZMAX= 47
IXMIN= -97 IYMIN= -51 IZMIN= -47


serial 3D FFT for wavefunctions
parallel 3D FFT for charge:
minimum data exchange during FFTs selected (reduces bandwidth)

Re: Incomplete simulation of single point SCF calculation

Posted: Mon May 09, 2022 7:25 am
by alex
Hello,

have you been patient enough? 400 atoms and 4 cores take a very long time even for finishing set-up. I'd guess several hours at least ...

Hth

alex

Re: Incomplete simulation of single point SCF calculation

Posted: Mon May 09, 2022 9:52 am
by aniket_singha
Thank you sir for the reply...
Made the changes as per your suggestions on no. of cores. Will update if anything such arises....