Hello,
I’m working on LAMMPS plus QE, and need a large input problem that can generate tens of GB memory footprint.
Can you give me a hint on how to generate such an input problem?
Thank you. I appreciate your help.
Zeshi Liu
Hello,
I’m working on LAMMPS plus QE, and need a large input problem that can generate tens of GB memory footprint.
Can you give me a hint on how to generate such an input problem?
this really isn’t a LAMMPS question, but a QE question, so you’d get better answers there.
to increase the memory needs (and i assume you care about QE, as it is by far the more memory hungry application) you can:
a) make a supercell
b) use norm-conserving pseudopotentials and increase the plane wave cutoff to a larger than usual number.
axel.
Dear Axel,
Thank you for your answer.
Can you give more information on how to make supercell and use norm-conserving pseudopotentials and increase the plane wave cutoff to a larger than usual number?
I am not a domain expert and lack such knowledge.
Thank you. I appreciate your help.
Zeshi
Dear Axel,
Thank you for your answer.
Can you give more information on how to make supercell and use norm-conserving pseudopotentials and increase the plane wave cutoff to a larger than usual number?
I am not a domain expert and lack such knowledge.
sorry to be blunt about this, but if you don’t know what you are doing, you should not try to run QM/MM calculations. those are really tricky and you do have to know what you are doing or else you are wasting your time.
if you are doing this on behalf of somebody else, you need to communicate with that somebody else. there is not enough space in an e-mail to explain to you everything you need to learn to be able to do what you want to do correctly.
at any rate, those are questions pertaining to the quantum espresso part and this is the LAMMPS mailing list, so you are asking in the wrong forum, too.
axel.