Page 3 of 3
Re: stuck in caluclation of static inverse dielectric matrix
Posted: Fri May 03, 2019 8:54 am
by Daniele Varsano
Dear Zhou,
your script seems to me to be correct.
Can you try to reduce the number of bands just as a test to be sure the problem is related to memory?
Best,
Daniele
Re: stuck in caluclation of static inverse dielectric matrix
Posted: Fri May 03, 2019 10:33 am
by ljzhou86
Daniele Varsano wrote:Dear Zhou,
your script seems to me to be correct.
Can you try to reduce the number of bands just as a test to be sure the problem is related to memory?
Best,
Daniele
Dear Daniele
I would like to share my experimence:
I have fixed this issue. It is actually the issue of memory. In my submission script, I am launching intel's mpirun . But the mpirun is not able to pick up the options I set with #SBATCH. I switched to srun as "srun -n 192 yambo -F ..." and found that I could use more large memory per core in this way.
Thanks
Re: stuck in caluclation of static inverse dielectric matrix
Posted: Fri May 03, 2019 10:36 am
by Daniele Varsano
Dear Zhou,
I'm glad to know the problem is solved.
Best,
Daniele