VASP 6.3.1 KPOINTS_OPT and NPAR

[chenghao_liu1]

Hi!

I'm encountering three issues when running with the new KPOINTS_OPT file for band structure calculations (GGA, PBE). Everything is normal if I run a separate SCF followed by a ICHARG=11 run. Compiled with intelmpi/2019.7.217, on Graham, a Canadian cluster.

1. The first error seems to relate to NPAR. When I set NCORE (or NPAR), after the SCF converges, I get an error as the following before KPOINTS_OPT. When NCORE is default, this goes away. My concern is that if I cannot use NCORE or NPAR, hybrid functional calculations may become very expensive.

Code: Select all

 -----------------------------------------------------------------------------
|                                                                             |
|     EEEEEEE  RRRRRR   RRRRRR   OOOOOOO  RRRRRR      ###     ###     ###     |
|     E        R     R  R     R  O     O  R     R     ###     ###     ###     |
|     E        R     R  R     R  O     O  R     R     ###     ###     ###     |
|     EEEEE    RRRRRR   RRRRRR   O     O  RRRRRR       #       #       #      |
|     E        R   R    R   R    O     O  R   R                               |
|     E        R    R   R    R   O     O  R    R      ###     ###     ###     |
|     EEEEEEE  R     R  R     R  OOOOOOO  R     R     ###     ###     ###     |
|                                                                             |
|     VASP internal routines have requested a change of the k-point set.      |
|     Unfortunately, this is only possible if NPAR=number of nodes.           |
|     Please remove the tag NPAR from the INCAR file and restart the          |
|     calculation.                                                            |
|                                                                             |
|       ---->  I REFUSE TO CONTINUE WITH THIS SICK JOB ... BYE!!! <----       |
|                                                                             |
 -----------------------------------------------------------------------------

2. The second error happens regardless what I use. I get a bunch of this 'LRF_COMMUTATOR internal error: the vector H(1)-e(1) S(1) |phi(0)> is not orthog onal to |phi(0)> (-4.895082546823611E-016,2.781571800288059E-002)' before 'Start KPOINTS_OPT (optional k-point list driver)'. Although, it does not prevent KPOINTS_OPT being used for further calculation and resulting in a 'normal' termination.

3. I don't see a PROCAR_OPT file so I can't get the band structure.

Thank you in advance!

My INCAR is as follows:
EDIFF = 1e-07
IVDW = 12
ADDGRID= True
PREC = Accurate
ICHARG = 2
ISMEAR = 0
SIGMA = 0.01
ENCUT = 500
LMAXMIX = 4
NELM = 300
LREAL = AUTO
EMIN = -6
NEDOS = 1201
EMAX = 3
PSTRESS = 0

My KPOINTS is as follows:
Automatic kpoint scheme
0
Gamma
3 3 13

My KPOINTS_OPT is as follows:
K-Path
20 ! Grid points
line-mode
reciprocal
0.0000000000 0.0000000000 0.0000000000 ! \Gamma
0.5000000000 0.0000000000 0.0000000000 ! M

0.5000000000 0.0000000000 0.0000000000 ! M
0.3333333333 0.3333333333 0.0000000000 ! K

0.3333333333 0.3333333333 0.0000000000 ! K
0.0000000000 0.0000000000 0.0000000000 ! \Gamma

0.0000000000 0.0000000000 0.0000000000 ! \Gamma
0.0000000000 0.0000000000 0.5000000000 ! A

0.0000000000 0.0000000000 0.5000000000 ! A
0.5000000000 0.0000000000 0.5000000000 ! L

0.5000000000 0.0000000000 0.5000000000 ! L
0.3333333333 0.3333333333 0.5000000000 ! H

0.3333333333 0.3333333333 0.5000000000 ! H
0.0000000000 0.0000000000 0.5000000000 ! A

0.5000000000 0.0000000000 0.5000000000 ! L
0.5000000000 0.0000000000 0.0000000000 ! M

0.3333333333 0.3333333333 0.0000000000 ! K
0.3333333333 0.3333333333 0.5000000000 ! H

[jonathan_lahnsteiner2]

Dear chenghao_liu1,

There was already a similar question in this thread
forum/viewtopic.php?f=4&t=18688 .
Maybe you can update your code and check if this
fixes your problem.

Otherwise could you please send me your POSCAR file and tell
me exactly what POTCAR file you used. Then we will check
what the issue is.

All the best Jonathan

[chenghao_liu1]

Hi Jonathan,

Thanks for your reply. I updated the code according to the thread below and recompiled, and still I have the same issues. Please find the zip file attached here which includes 2 runs, one with ncore=32 (fails) and one without (finishes but with LRF_COMMUTATOR internal error and without PROCAR_OPT), along with the .eb file I used to compile VASP. Thanks in advance.

Kind regards,
Chenghao

[jonathan_lahnsteiner2]

Dear chenghao_liu1,

Right now it is not possible to use the KPOINTS_OPT file together with the INCAR tags NPAR or NCORE.

| VASP internal routines have requested a change of the k-point set. |
| Unfortunately, this is only possible if NPAR=number of nodes. |
| Please remove the tag NPAR from the INCAR file and restart the |
| calculation.

This is also what the error message tries to hint you to.
I hope this solves your issue.

All the best Jonathan

[chenghao_liu1]

Hi Jonathan,

Thanks for your reply. Indeed that is one of the issues. But even when I run without NCORE/NPAR settings, I still get the LRF_COMMUTATOR internal error, and I do not get a PROCAR_OPT file... Could you let me know if I'm doing something wrong or this is a bug? Thanks!

Kind regards,
Chenghao

[jonathan_lahnsteiner2]

Dear Chenghao,

I was checking your INCAR file again. You have to set LORBIT=11 otherwise
no PROCAR_OPT file will be written. You can ignore the
LRF_COMMUTATOR internal error.

All the best Jonathan

[chenghao_liu1]

Thank you Jonathan, I appreciate your reply!

Kind regards,
Chenghao