[Pw_forum] Fwd: Error in final scf calculation using vc-relax

[Malte Sachs]

Dear all,

I am performing spin polarized vc-relax calculations using QE 6.1 and 
6.2. For some inputs I am getting an error message in the final 
scf-calculation after reaching the final coordinates that freeze the 
calculation:

...

      Initial potential from superposition of free atoms
      Starting wfc are  120 randomized atomic wfcs +    2 random wfc

*Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.*

...

The error is reproducable for QE 6.1 and QE 6.2. However, using QE 
version 6.0 with the same compilation parameters the calculation works 
fine. I have noticed some differences in the output comparing version 
6.0 and 6.1/6.2. In version 6.0 the starting magnetic structure for the 
final scf calculation is initialized as provided by me in the input-file 
with the parameter "starting_magnetization":

/  Starting magnetic structure//
//     atomic species   magnetization//
//        U            0.010//
//        Rh           0.010/

However, in version 6.1/6.2 I get:

/ Starting magnetic structure//
//     atomic species   magnetization//
//        U              NaN//
//        Rh           0.000/

Did the initialization of the final scf step change from version 6.0 to 
6.1? And could this lead to the error? I am just a beginner in 
compilation issues. So I really would appreciate, if someone could give 
me a hint, why this error occurs. Attached you can find the input file. 
If you wish, I can send you the two output files for the 6.0 and 6.1 
run. (Don't wonder: the structure doesn't contain Uranium atoms as the 
input-file is generated by a script for a variable composition search.)

Best regards,

Malte

-- 
Malte Sachs
Anorganische Chemie, Fluorchemie
Philipps-Universität Marburg
Hans-Meerwein-Straße 4
35032 Marburg (Paketpost: 35043 Marburg)
Tel.: +49 (0)6421 28 - 25 39 5
http://www.uni-marburg.de/fb15/ag-kraus/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20171030/6ac94eae/attachment.html>
-------------- next part --------------
&CONTROL
  calculation  = 'vc-relax'
  restart_mode = 'from_scratch'
  prefix       = 'Rh_U'
  pseudo_dir = '/home/Sachsm/Promotion/Pseudos'
  wf_collect   = .false.
  disk_io = 'none'
  etot_conv_thr=1.0D-4
  forc_conv_thr=1.0D-3
  nstep = 30
/
&SYSTEM
  ibrav     = 0 
  nat       = 12
  ntyp      = 2
  ecutwfc   = 95
  ecutrho   = 800 
  nspin     = 2
  starting_magnetization(1) = 1.D-2
  starting_magnetization(2) = 1.D-2
  nosym     = .true.
  occupations = 'smearing'
  smearing = 'gaussian'
  degauss = 0.02
/
&ELECTRONS
  conv_thr = 5.D-7
  mixing_beta = 0.3
  mixing_mode = 'local-TF'
  electron_maxstep = 100
  mixing_ndim = 16
  !tqr = .true.
/
&IONS
/
&CELL
  cell_factor = 2.0
/
ATOMIC_SPECIES
 U 238.03 U.pbe-spn-kjpaw_MS.UPF
 Rh 102.91 Rh.pbe-spn-kjpaw_psl.1.0.0.UPF

 
CELL_PARAMETERS bohr
  9.7842   0.0000   0.0000
 -0.0002   8.9386   0.0000
  1.6341  -0.0019  13.1389
ATOMIC_POSITIONS {crystal} 
  Rh       0.668866     0.665740     0.084843
  Rh       0.335317     0.332555     0.918174
  Rh       0.835573     0.832321     0.418172
  Rh       0.168838     0.165944     0.584877
  Rh       0.835353     0.165795     0.918168
  Rh       0.168809     0.832431     0.084758
  Rh       0.668806     0.332215     0.584791
  Rh       0.335521     0.665993     0.418274
  Rh       0.002294     0.332422     0.251497
  Rh       0.501996     0.832334     0.751464
  Rh       0.502300     0.165962     0.251510
  Rh       0.001945     0.665799     0.751470
K_POINTS {automatic} 
   4    5    3  0 0 0
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5303 bytes
Desc: S/MIME Cryptographic Signature
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20171030/6ac94eae/attachment.p7s>