[QE-users] The routines to calculate Hubbard forces are very slow

[Timrov Iurii]

Dear Julien,

Your system has 96 atoms, so indeed the Hubbard forces (and stress) calculations with "ortho-atomic" Hubbard projectors are expected to be very slow. Switching to "atomic" will make it much faster. However, the value of U and/or V changes when you change the projectors, and the results might also change substantially when you change the projectors.

HTH

Iurii

----------------------------------------------------------
Dr. Iurii TIMROV
Tenure-track scientist
Laboratory for Materials Simulations (LMS)
Paul Scherrer Institut (PSI)
CH-5232 Villigen, Switzerland
+41 56 310 62 14
https://www.psi.ch/en/lms/people/iurii-timrov
________________________________
From: users <users-bounces at lists.quantum-espresso.org> on behalf of julien_barbaud at sjtu.edu.cn <julien_barbaud at sjtu.edu.cn>
Sent: Monday, January 13, 2025 19:31
To: 'Quantum ESPRESSO users Forum' <users at lists.quantum-espresso.org>
Subject: [QE-users] The routines to calculate Hubbard forces are very slow


Dear users,



I have been using Hubbard calculations recently in the context of polaron studies (I am using DFT+V on 2 atoms to study a polaronic dimer).

I did not encounter any particular issue while tuning the V value using the generalized Koopman’s condition. However, when trying to compute forces with DFT+V, the calculations appear to be very slow: on a 36 minute run, 28 were spent on the force_hub routine. Is this supposed to happen? This seems odd, and I suspect I might be doing something wrong. Other calculations run smoothly, it is only the forces that seem to slow everything down

I included the input and output files for an example calculation, as well as the slurm file used to control the parallelization parameters.



Thanks in advance,

Julien
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20250114/da97c46c/attachment.html>