[Pw_forum] pw.x stops when the calculation 'bands' reaches a specific kpoint

[stefano de gironcoli]

cholesky decomposition is used in Davidson diagonalization
it's a problem related to the fact that the matrix is not well formed 
due to some reason.
smearing type, mixinng mode, degauss etc apply to the scf cycle and 
should be largely un important at the band calculation stage.

you can try to change diagonalization scheme, using conjugate gradient 
that does not call the cholesky decomposition routine.

stefano

On 23/01/2016 11:57, Mohamad Moadeli wrote:
> Dear all,
> I am trying to calculate the band structure of NI(111). The 'scf' 
> calculation ran fine, but the 'bands' one stoped at a specific kpoint. 
> Based on the 5.6 part of FAQ, we changed the smearing (gaussian, mp, 
> mv), mixing mode (plain, local-TF), dgauss (0.01, 0.001), kpoint mesh 
> (300, 1000) and....; but the problem did not solved.
>
> The input file:
> ==========================================================
> &control
>     calculation= 'bands',
>     restart_mode= 'from_scratch' ,
>     pseudo_dir = '/home/pseudo/' ,
>     outdir= 'tmp' ,
>     prefix= 'lda-fm'
>     verbosity='high' ,
>     tstress= .TRUE.
>     tprnfor= .TRUE.
> /
> &system
>     ibrav = 4,
>     celldm(1)= 4.639055, celldm(3)= 12.00000,
>     nat =6 , ntyp =1,
>     nspin =2, starting_magnetization(1)=0.7,
>     ecutwfc = 65, ecutrho = 450,
>     occupations = 'smearing',smearing='gaussian',degauss=0.001,
>     nosym=.true. ,
> /
> &electrons
>     conv_thr = 1.0d-8,
>     mixing_mode = 'plain',
>     mixing_beta = 0.3 ,
>     diagonalization = 'david',
>     electron_maxstep =400,
> /
> ATOMIC_SPECIES
>     Ni   58.6934   Ni.pz-n-rrkjus_psl.0.1.UPF
>
> ATOMIC_POSITIONS (crystal)
> Ni       0.000000000   0.000000000   0.000000000    0   0   0
> Ni       0.666666794   0.333333397   0.066178769
> Ni       0.333333588   0.666666794   0.133051982
> Ni       0.000000000   0.000000000   0.199970765
> Ni       0.666666794   0.333333397   0.266756139
> Ni       0.333333588   0.666666794   0.333562723
>
> K_POINTS crystal
>         1000
>         0.0000000000    0.0000000000    0.0000000000    1.0
>         0.0007898815    0.0007898815    0.0000000000    1.0
>         0.0015797630    0.0015797630    0.0000000000    1.0
>         0.0023696445    0.0023696445    0.0000000000    1.0
>         0.0031595261    0.0031595261    0.0000000000    1.0
>         0.0039494076    0.0039494076    0.0000000000    1.0
>         0.0047392891    0.0047392891    0.0000000000    1.0
>         0.0055291706    0.0055291706    0.0000000000    1.0
>         0.0063190521    0.0063190521    0.0000000000    1.0
>         0.0071089336    0.0071089336    0.0000000000    1.0
> 0.0078988152    0.0078988152    0.0000000000    1.0
>         0.0086886967    0.0086886967    0.0000000000    1.0
>         0.0094785782    0.0094785782    0.0000000000    1.0
> ........
> ==========================================================
>
>
> The output file:
> ==========================================================
> ... .
> ......
>      Band Structure Calculation
>      Davidson diagonalization with overlap
>
>      Computing kpt #:     1
>      total cpu time spent up to now is       12.4 secs
>
>      Computing kpt #:     2
>      total cpu time spent up to now is       15.7 secs
>
>      Computing kpt #:     3
>      total cpu time spent up to now is       19.4 secs
>
>      Computing kpt #:     4
>      total cpu time spent up to now is       24.2 secs
>
>      Computing kpt #:     5
>      total cpu time spent up to now is       27.9 secs
>
>      Computing kpt #:     6
>      total cpu time spent up to now is       31.7 secs
>
>      Computing kpt #:     7
>      total cpu time spent up to now is       37.1 secs
>
>      Computing kpt #:     8
>      total cpu time spent up to now is       42.0 secs
>
>      Computing kpt #:     9
>      total cpu time spent up to now is       47.7 secs
>
>      Computing kpt #:    10
>      total cpu time spent up to now is       53.0 secs
>
>      Computing kpt #:    11
>
>
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>      Error in routine  cdiaghg (130):
>       problems computing cholesky
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
>      stopping ...
> .......
> ==========================================================
> We changed the number of kpoints, but when it reached /that/ kpoint, 
> it stoped.
>
>
> Any help would be greatly appreciated.
>
> Mohammad Moaddeli,
>
> Chahid Chamran University of Ahvaz, Ahvaz.
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160124/a1d1ab37/attachment.html>