[Pw_forum] HSE calculation stops progressing after convergence
Giuseppe Mattioli
giuseppe.mattioli at ism.cnr.it
Fri Feb 16 12:46:53 CET 2018
Dear Shoaib Muhammad
First of all EXX calculations are *far more stable* when you use norm
conserving pseudopotentials. ultrasoft pseudopotentials such as yours
do not even guarantee particular speedup: the bottleneck is on density
cutoff rather than on wavefunction cutoff.
Moreover, EXX calculations are generally more expensive than GGA, even
at Gamma (and your cell is small and would require a proper k-points
mesh). As you did not post further details, I can only suggest that you:
1) check that the memory requirements for your calculation are
compatible with your actual resources; an (over)estimate of the total
memory is printed at the beginning of the output, before the beginning
of scf iterations.
2) use a recent version of the code (6.2 or newer), which
automatically implement the ACE algorithm which speedups the calculation
3) play a bit with ecutfock (ecutfock=2*ecutrho is generally safe,
even less sometimes) which saves a lot of memory
HTH
Giuseppe
Quoting Shoaib <mshoaibce at gmail.com>:
> Dear Members,
>
>
>
> I am trying a simple HSE calculation but it stop progressing after some time
> where output file says "convergence has been achieved in 10 iterations"
> without showing any further error. CPU usage still remains high but nothing
> is further printed in the output file even if I wait for hours. Here is my
> input file:
>
>
>
>
>
> &CONTROL
>
> calculation = "scf"
>
> max_seconds = 8.64000e+04
>
> outdir = "./"
>
> prefix = "espresso"
>
> pseudo_dir = " \.pseudopot"
>
> title = "LiFePO4 (SCF)"
>
> wf_collect = .TRUE.
>
> wfcdir = "./"
>
> tstress = .TRUE.
>
> tprnfor = .true.
>
> verbosity = 'high'
>
> /
>
>
>
> &SYSTEM
>
> a = 4.75196e+00
>
> b = 6.08735e+00
>
> c = 1.04507e+01
>
> degauss = 1.00000e-02
>
> ecutrho = 3.50000e+02
>
> ecutwfc = 3.50000e+01
>
> exxdiv_treatment = "gygi-baldereschi"
>
> ibrav = 8
>
> input_dft = "hse"
>
> nat = 28
>
> nbnd = 132
>
> nqx1 = 1
>
> nqx2 = 1
>
> nqx3 = 1
>
> exx_fraction = 0.25
>
> ntyp = 4
>
> occupations = "smearing"
>
> smearing = "gaussian"
>
> x_gamma_extrapolation = .TRUE.
>
> /
>
>
>
> &ELECTRONS
>
> conv_thr = 1.00000e-06
>
> electron_maxstep = 200
>
> adaptive_thr = .TRUE.
>
> mixing_beta = 4.00000e-01
>
> startingpot = "atomic"
>
> startingwfc = "atomic+random"
>
> /
>
>
>
> K_POINTS {gamma}
>
> ATOMIC_SPECIES
>
> Li 6.94100 Li.pbe-n-van.UPF
>
> Fe 55.84500 Fe.pbe-sp-van.UPF
>
> P 30.97376 P.pbe-n-van.UPF
>
> O 15.99940 O.pbe-rrkjus.UPF
>
>
>
> ATOMIC_POSITIONS {angstrom}
>
> Li 0.000000 0.000000 0.000000
>
> Li 2.375978 3.043675 5.225334
>
> Li 2.375978 0.000000 5.225334
>
> Li 0.000000 3.043675 0.000000
>
> Fe 2.263451 4.565513 8.168085
>
> Fe 4.639429 1.521838 7.507916
>
> Fe 0.112526 4.565513 2.942751
>
> Fe 2.488504 1.521838 2.282583
>
> P 2.760801 4.565513 0.989699
>
> P 0.384823 1.521838 4.235635
>
> P 4.367133 4.565513 6.215033
>
> P 1.991155 1.521838 9.460968
>
> O 1.223358 4.565513 1.010109
>
> O 1.359820 2.756426 8.719891
>
> O 1.359820 0.287250 8.719891
>
> O 3.528598 1.521838 9.440558
>
> O 1.396395 1.521838 0.446421
>
> O 3.599335 1.521838 4.215225
>
> O 3.392136 5.800101 1.730777
>
> O 3.735797 3.330925 6.956111
>
> O 1.016158 2.756426 3.494557
>
> O 1.016158 0.287250 3.494557
>
> O 3.355560 4.565513 10.004246
>
> O 3.772373 4.565513 4.778913
>
> O 0.979582 1.521838 5.671755
>
> O 1.152620 4.565513 6.235443
>
> O 3.735797 5.800101 6.956111
>
> O 3.392136 3.330925 1.730777
>
>
>
>
>
> And last few lines of output file are
>
> End of self-consistent calculation
>
>
>
> k = 0.0000 0.0000 0.0000 ( 3575 PWs) bands (ev):
>
>
>
> -79.0030 -79.0023 -79.0019 -78.9839 -45.2255 -45.2252 -45.2234 -45.2059
>
> -45.1487 -45.1485 -45.1473 -45.1309 -44.9335 -44.9334 -44.9323 -44.9147
>
> -15.3106 -15.1682 -15.0711 -14.9777 -12.6234 -12.6211 -12.5757 -12.5494
>
> -12.5403 -12.5249 -12.5217 -12.4742 -12.4125 -12.3868 -12.3624 -12.3553
>
> -2.5810 -2.1773 -2.1375 -2.0024 -0.9665 -0.6145 -0.5687 -0.4843
>
> -0.1764 -0.1015 0.1339 0.2764 0.3243 0.3686 0.4349 0.4662
>
> 1.5656 1.9325 2.0392 2.0667 2.2272 2.2489 2.2587 2.2709
>
> 2.3867 2.5635 2.7745 2.9206 3.0322 3.0398 3.1022 3.1224
>
> 3.1659 3.1803 3.3798 3.4103 3.4174 3.5710 3.5829 3.5833
>
> 3.7764 3.8840 4.0086 4.0279 4.1655 4.3439 4.7040 4.8756
>
> 7.1039 7.1654 7.2249 7.2788 7.2957 7.3083 7.3196 7.3530
>
> 7.4510 7.4565 7.4624 7.4743 8.0415 8.0544 8.0944 8.3315
>
> 8.4876 8.5099 8.6253 8.6437 10.2682 12.0817 12.4471 13.1204
>
> 13.3931 13.9195 13.9632 13.9830 14.1605 14.3201 14.5312 14.7573
>
> 14.8969 14.9143 15.1052 15.6954 15.7497 16.0537 16.3593 16.5595
>
> 16.5733 16.6684 16.6782 16.7533 16.9513 17.0567 17.2068 17.6553
>
> 17.7728 17.7858 18.0596 18.5210
>
>
>
> occupation numbers
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
>
> 0.9994 0.9993 0.9991 0.9987 0.0019 0.0013 0.0003 0.0000
>
> 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
>
> 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
>
> 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
>
> 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000
>
> 0.0000 0.0000 0.0000 0.0000
>
>
>
> the Fermi energy is 7.7636 ev
>
>
>
> ! total energy = -1586.80562557 Ry
>
> Harris-Foulkes estimate = -1586.80581500 Ry
>
> estimated scf accuracy < 0.00024789 Ry
>
>
>
> convergence has been achieved in 10 iterations
>
>
>
>
>
> Shoaib Muhammad, PhD
>
> Department of Energy Science
>
> Sungkyunkwan University
>
> South Korea
GIUSEPPE MATTIOLI
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
Via Salaria Km 29,300 - C.P. 10
I-00015 - Monterotondo Scalo (RM)
Mob (*preferred*) +39 373 7305625
Tel + 39 06 90672342 - Fax +39 06 90672316
E-mail: <giuseppe.mattioli at ism.cnr.it>
More information about the users
mailing list