Molecules in Abinit

sponce · Post by **sponce** » Wed Dec 02, 2020 5:06 pm

Hello all,

I have a question linked with the calculation of molecules in Abinit.
I used the tbase1_3.in input as test case but instead of H2, I'm doing NH4.

The input file is as follow (The psp use the one of Psps_for_tests):

Code: Select all

ionmov  2
ntime   10
tolmxf  5.0d-4
toldff  5.0d-5
ntypat 2
znucl 7 1
natom 4
typat 1 2 2 2
 xcart
  0.0000000000    0.0000000000    0.0000000000
  -1.88972687777  0.00000000  -0.75589075111
  0.94486343888  1.63655147522  -0.75589075111
  0.94486343888  -1.63655147522 -0.75589075111
kptopt 0
nkpt 1
nstep 10
diemac 2.0
 pp_dirpath "$ABI_PSPDIR"
 pseudos "Pseudodojo_nc_sr_04_pbe_standard_psp8/N.psp8, PseudosGTH_pwteter/01h.pspgth "
acell 30 30 30
ecut 20.0

The following input crash (segfault) at the level of

Code: Select all

---SELF-CONSISTENT-FIELD CONVERGENCE--------------------------------------------
 
 getcut: wavevector=  0.0000  0.0000  0.0000  ngfft= 125 125 125
         ecut(hartree)=     20.000   => boxcut(ratio)=   2.05315
 
 1/G**2 cut-off applied in the following step : cutoff-mode = CRYSTAL
 
 1/G**2 cut-off applied in the following step : cutoff-mode = CRYSTAL
 
 1/G**2 cut-off applied in the following step : cutoff-mode = CRYSTAL

===================================================================================
=   BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
=   RANK 0 PID 693502 RUNNING AT theossrv8
=   KILLED BY SIGNAL: 11 (Segmentation fault)
===================================================================================

However, If I reduce the ecut or decrease the box size, then it works.
For example making

Code: Select all

acell 30 30 30
ecut 10.0

runs and goes to the end.

I'm not sure I understand why I cannot use a large ecut when the cell is big ?

One should be able to systematically increase the ecut.

Many thanks,
Samuel

gmatteo · Post by **gmatteo** » Wed Dec 02, 2020 9:51 pm

Hi Samuel,

the input file works in sequential on my laptop.
The max memory allocated on the heap is ~700 Mb that is consistent with the value reported by Abinit at the beginning.

Perhaps it's a problem with the stack triggered by automatic arrays.
Did you try to increase the stack size limit with e.g.: `ulimit -s unlimited` ?

M

sponce · Post by **sponce** » Thu Dec 03, 2020 10:06 am

Yep, that's working !

Thanks Matteo !

ABINIT Discussion Forums

Molecules in Abinit [SOLVED]

Molecules in Abinit [SOLVED]

Re: Molecules in Abinit

Re: Molecules in Abinit