Slow convergence with hybrid functional

[KieranSong]

Dear All,

I used hybrid functional (HSE06) to calculate the band structure but the convergence of the calculation was very slow. After 72 hours, the energy difference (dE) still oscillated around 0.001 eV. There are only 24 C atoms in the system. Would anyone please give me some suggestions on converging the calculation faster? Thank you.

Here is the INCAR file

System = new
ISTART = 1
LWAVE = .FALSE.
LCHARGE = .TRUE.
ICHARG = 0
IVDW = 12
ISPIN = 2
MAGMOM = 0.043 -0.042 -0.019 0.019 -0.043 0.042 -0.043 0.042 0.019\
-0.019 0.043 -0.042 -0.043 0.042 0.019 -0.019 0.043 -0.042\
0.043 -0.042 -0.019 0.019 -0.043 0.042
LORBIT = 11
ENCUT = 500
ISMEAR = 0
SIGMA = 0.03
PREC = Accurate
LHFCALC = .TRUE.
AEXX = 0.25
HFSCREEN = 0.2
ALGO = Damped
TIME = 0.4
PRECFOCK = Fast
ISYM = 3
LDIAG = .TRUE.
LREAL = Auto
NELMIN = 5
NELM = 700
NELMDL = -10
EDIFF = 1e-5
NPAR = 32
KPAR= 16

Here is the output file

The job 22442524 is running on nid0[5782-5787,5789-5798]
running on 256 total cores
distrk: each k-point on 16 cores, 16 groups
distr: one band on 1 cores, 16 groups
using from now: INCAR
vasp.5.4.4.18Apr17-6-g9f103f2a35 (build Oct 30 2019 15:40:33) complex

POSCAR found type information on POSCAR C
POSCAR found : 1 types and 24 ions
scaLAPACK will be used

-----------------------------------------------------------------------------
| |
| W W AA RRRRR N N II N N GGGG !!! |
| W W A A R R NN N II NN N G G !!! |
| W W A A R R N N N II N N N G !!! |
| W WW W AAAAAA RRRRR N N N II N N N G GGG ! |
| WW WW A A R R N NN II N NN G G |
| W W A A R R N N II N N GGGG !!! |
| |
| For optimal performance we recommend to set |
| NCORE= 4 - approx SQRT( number of cores) |
| NCORE specifies how many cores store one orbital (NPAR=cpu/NCORE). |
| This setting can greatly improve the performance of VASP for DFT. |
| The default, NCORE=1 might be grossly inefficient |
| on modern multi-core architectures or massively parallel machines. |
| Do your own testing !!!! |
| Unfortunately you need to use the default for GW and RPA calculations. |
| (for HF NCORE is supported but not extensively tested yet) |
| |
-----------------------------------------------------------------------------

-----------------------------------------------------------------------------
| |
| W W AA RRRRR N N II N N GGGG !!! |
| W W A A R R NN N II NN N G G !!! |
| W W A A R R N N N II N N N G !!! |
| W WW W AAAAAA RRRRR N N N II N N N G GGG ! |
| WW WW A A R R N NN II N NN G G |
| W W A A R R N N II N N GGGG !!! |
| |
| You have enabled k-point parallelism (KPAR>1). |
| This developmental code was originally written by Paul Kent at ORNL, |
| and carefully double checked in Vienna. |
| GW as well as linear response parallelism added by Martijn Marsman |
| and Georg Kresse. |
| Carefully verify results versus KPAR=1. |
| Report problems to Paul Kent and Vienna. |
| |
-----------------------------------------------------------------------------

LDA part: xc-table for Pade appr. of Perdew

-----------------------------------------------------------------------------
| |
| W W AA RRRRR N N II N N GGGG !!! |
| W W A A R R NN N II NN N G G !!! |
| W W A A R R N N N II N N N G !!! |
| W WW W AAAAAA RRRRR N N N II N N N G GGG ! |
| WW WW A A R R N NN II N NN G G |
| W W A A R R N N II N N GGGG !!! |
| |
| Your generating k-point grid is not commensurate to the symmetry |
| of the lattice. This can cause slow convergence with respect |
| to k-points for HF type calculations |
| suggested SOLUTIONS: |
| ) if not already the case, use automatic k-point generation |
| ) shift your grid to Gamma (G) (e.g. required for hex or fcc lattice) |
| |
-----------------------------------------------------------------------------

found WAVECAR, reading the header
number of k-points has changed, file: 61 present: 561
trying to continue reading WAVECAR, but it might fail
POSCAR, INCAR and KPOINTS ok, starting setup
FFT: planning ...
reading WAVECAR
the WAVECAR file was read successfully
initial charge from wavefunction
entering main loop
N E dE d eps ncg rms ort
DAV: 1 -0.310449505422E+02 -0.31045E+02 -0.24020E+01 35904 0.142E+01
DAV: 2 -0.314504830256E+02 -0.40553E+00 -0.40548E+00 35904 0.892E+00
DAV: 3 -0.315170568035E+02 -0.66574E-01 -0.66564E-01 35904 0.353E+00
DAV: 4 -0.315331448015E+02 -0.16088E-01 -0.16086E-01 35904 0.139E+00
DAV: 5 -0.315377703876E+02 -0.46256E-02 -0.46250E-02 35904 0.741E-01
DAV: 6 -0.315392413315E+02 -0.14709E-02 -0.14708E-02 35904 0.355E-01
DAV: 7 -0.315397497945E+02 -0.50846E-03 -0.50842E-03 35904 0.207E-01
DAV: 8 -0.315399393488E+02 -0.18955E-03 -0.18954E-03 35904 0.108E-01
DAV: 9 -0.315400155056E+02 -0.76157E-04 -0.76151E-04 35904 0.670E-02
DAV: 10 -0.315400482090E+02 -0.32703E-04 -0.32701E-04 35904 0.382E-02
gam= 0.000 g(H,U,f)= 0.210E+02 0.488E+00 0.212E-01 ort(H,U,f) = 0.000E+00 0.000E+00 0.000E+00
SDA: 11 -0.237406440876E+03 -0.20587E+03 -0.86126E+01 35904 0.215E+02 0.000E+00
gam= 0.382 g(H,U,f)= 0.216E+01 0.146E+00 0.747E-02 ort(H,U,f) =-0.118E+01 0.234E+00 0.131E-01
...
...
...
DMP: 71 -0.242430035071E+03 -0.32650E-03 -0.44451E-03 35904 0.916E-03 0.511E-03
gam= 0.382 g(H,U,f)= 0.847E-03 0.934E-04 0.953E-08 ort(H,U,f) = 0.261E-03 0.163E-03 0.167E-07
DMP: 72 -0.242430342133E+03 -0.30706E-03 -0.44090E-03 35904 0.940E-03 0.424E-03
gam= 0.382 g(H,U,f)= 0.937E-03 0.853E-04 0.160E-07 ort(H,U,f) = 0.169E-03 0.149E-03 0.280E-07
DMP: 73 -0.242430626170E+03 -0.28404E-03 -0.45746E-03 35904 0.102E-02 0.318E-03
gam= 0.382 g(H,U,f)= 0.110E-02 0.781E-04 0.269E-07 ort(H,U,f) = 0.922E-05 0.136E-03 0.469E-07
DMP: 74 -0.242430883927E+03 -0.25776E-03 -0.49475E-03 35904 0.118E-02 0.145E-03
gam= 0.382 g(H,U,f)= 0.121E-02 0.714E-04 0.436E-07 ort(H,U,f) =-0.142E-03 0.124E-03 0.761E-07
DMP: 75 -0.242431127732E+03 -0.24381E-03 -0.51132E-03 35904 0.129E-02-0.178E-04
gam= 0.382 g(H,U,f)= 0.124E-02 0.653E-04 0.683E-07 ort(H,U,f) =-0.227E-03 0.114E-03 0.119E-06
DMP: 76 -0.242431360702E+03 -0.23297E-03 -0.50605E-03 35904 0.131E-02-0.113E-03

Any suggestion appreciated.

Kind regards,

Kieran

[marie-therese.huebsch]

Hi Kieran,

as discussed in the other post , the main issue is that your calculation is not performed using the proper k mesh. That is why you are receiving the following warning:

-----------------------------------------------------------------------------
| |
| W W AA RRRRR N N II N N GGGG !!! |
| W W A A R R NN N II NN N G G !!! |
| W W A A R R N N N II N N N G !!! |
| W WW W AAAAAA RRRRR N N N II N N N G GGG ! |
| WW WW A A R R N NN II N NN G G |
| W W A A R R N N II N N GGGG !!! |
| |
| Your generating k-point grid is not commensurate to the symmetry |
| of the lattice. This can cause slow convergence with respect |
| to k-points for HF type calculations |
| suggested SOLUTIONS: |
| ) if not already the case, use automatic k-point generation |
| ) shift your grid to Gamma (G) (e.g. required for hex or fcc lattice) |
| |
-----------------------------------------------------------------------------

Additionally, the other warnings actually contain advice on how to speed up your calculation. Did you try to follow them?

Best regards,
Marie-Therese