Dear Jeames,
the fact that tetraheadra are used in the nscf run before dos.x seems to
resonate with a bug that has been pointed out recently:
https://www.mail-archive.com/[email protected]/msg32918.html
that is mostly related to the use of QEXML fmt in 6.0 and 6.1.
QE-6.2, using by default a different format (unless -D__OLDXML is
specified), is unaffected...
take care
Andrea
input that can be run in a reasonable time and an exact description of
compilation options (serial, parallel MPI,
parallel MPI+OpenMP) and kind of execution (number of processors etc.)
Good day Paolo,
Okay, so here are the input files for the tests I did on parallel MPI+OpenMP.
Please let me know if you need anything else.
Parallel version (MPI & OpenMP), running on 12 processor cores
Number of MPI processes: 4
Threads/MPI process: 3
scf.in
&CONTROL
calculation='scf',
outdir='silicon',
prefix='calc',
pseudo_dir='.',
verbosity='low',
tstress=.false.,
tprnfor=.false.,
/
&SYSTEM
ibrav=2,
celldm(1)=10.2623466921d0,
nat=2,
ntyp=1,
ecutwfc=30.0d0,
ecutrho=300.0d0,
nbnd=10,
input_dft='PBE',
/
&ELECTRONS
diagonalization='david',
conv_thr=1d-08,
mixing_mode='plain',
mixing_beta=0.700d0,
/
ATOMIC_SPECIES
Si 28.085500d0 Si.pbe-n-rrkjus_psl.0.1.UPF
ATOMIC_POSITIONS {alat}
Si 0.0000000000d0 0.0000000000d0 0.0000000000d0
Si 0.2500000000d0 0.2500000000d0 0.2500000000d0
K_POINTS {automatic}
9 9 9 0 0 0
nscf.in
&CONTROL
calculation='nscf',
outdir='silicon',
prefix='calc',
pseudo_dir='.',
verbosity='low',
tstress=.false.,
tprnfor=.false.,
/
&SYSTEM
ibrav=2,
celldm(1)=10.2623466921d0,
nat=2,
ntyp=1,
ecutwfc=30.0d0,
ecutrho=300.0d0,
nbnd=10,
input_dft='PBE',
occupations='tetrahedra',
/
&ELECTRONS
diagonalization='david',
conv_thr=1d-08,
mixing_mode='plain',
mixing_beta=0.700d0,
/
ATOMIC_SPECIES
Si 28.085500d0 Si.pbe-n-rrkjus_psl.0.1.UPF
ATOMIC_POSITIONS {alat}
Si 0.0000000000d0 0.0000000000d0 0.0000000000d0
Si 0.2500000000d0 0.2500000000d0 0.2500000000d0
K_POINTS {automatic}
32 32 32 0 0 0
dos.in
&DOS
outdir='silicon',
prefix='calc',
Emin=-10,
Emax=20,
DeltaE=0.05,
fildos='silicon.dos',
/
Best regards,
James
2017-11-13 6:24 GMT+09:00 Paolo Giannozzi <[email protected]>:
On Sun, Nov 12, 2017 at 6:41 AM, Wilbert James Futalan
<[email protected]> wrote:
this problem really gets on my nerves
reports of problems without needed information get on MY nerves. Please provide
an input that can be run in a reasonable
time and an exact description of compilation options (serial, parallel MPI,
parallel MPI+OpenMP) and kind of execution
(number of processors etc.) that lead to such problem
--
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
--
Engr. Wilbert James C. Futalan
Research Fellow I
Laboratory of Electrochemical EngineeringDepartment of Chemical Engineering
University of the Philippines - Diliman
--
Andrea Ferretti, PhD
S3 Center, Istituto Nanoscienze, CNR
via Campi 213/A, 41125, Modena, Italy
Tel: +39 059 2055322; Skype: andrea_ferretti
URL: http://www.nano.cnr.it
_______________________________________________
Pw_forum mailing list
[email protected]
http://pwscf.org/mailman/listinfo/pw_forum
