Convergence Problem

[iduygnay]

Details

This calculation is really important for me, I tried different ways but still could not converge. Can anyone help me to check my INPUT? dft.zip

Have you read FAQ on the online manual http://abacus.deepmodeling.com/en/latest/community/faq.html

• [X] Yes, I have read the FAQ part on online manual.

Task list for Issue attackers (only for developers)

• [ ] Understand the problem or question described by the user.
• [ ] Check if the issue is a known problem or has been addressed in the documentation.
• [ ] Test the issue or problem on a similar system or environment, if possible.
• [ ] Identify the root cause or provide clarification on the user's question.
• [ ] Provide a step-by-step guide, including any necessary resources, to resolve the issue or answer the question.
• [ ] If the issue is related to documentation, update the documentation to prevent future confusion (optional).
• [ ] If the issue is related to code, consider implementing a fix or improvement (optional).
• [ ] Review and incorporate any relevant feedback from users or developers.
• [ ] Ensure the user's issue is resolved or their question is answered and close the ticket.

[WHUweiqingzhou]

@iduygnay have you tried this notebook?

[iduygnay]

I tried to adjust mix_gg0 to 1.5, mix_beta to 0.6 and some other values, degauss to 0.01 and 0.02, smearing method to fixed, but all of them always converge to 1E-3 and can not be smaller.

[WHUweiqingzhou]

I notice it is a SOC calculation, I will suggest a quite small mixing_beta , like 0.2 or 0.1

[iduygnay]

It still could not converge.

[dyzheng]

I suggest to set "noncolin false" when there is no initial atomic-mag is setting, it usually can improve convergence.

[iduygnay]

I set noncolin as default (which is false), still not converge.

[iduygnay]

I've used QE and solved this, QE can converge quite quickly. But it requires a large memory.

[iduygnay]

I tried symmetry = -1, mixing_beta = 0.01, soc lambda = 0.1, still not converge.

[iduygnay]

Even 32 atoms cannot converge.

[iduygnay]

Interesting! I found that abacus3.4.3 can converge, for the 32 atoms case with the same INPUT.

[iduygnay]

240 atoms case can also converge successfully! I think my problem is solved, if there is no necessary to keep this issue I can close it.

[WHUweiqingzhou]

@iduygnay Could you offer your input&output of 32 atoms. I believe this is caused by some numerical problem, and I will try to figure it out.

[iduygnay]

I hope to send it via WeChat since the file is too big.