[Pw_forum] segmentation fault
Ari P Seitsonen
Ari.P.Seitsonen at iki.fi
Sat Jun 20 19:30:59 CEST 2015
Dear Vishal Gupta,
Are you sure about your cell parametres? First dimension looks quite
large (looks more like 200-400 cores), the third one very small. The
number of k points is totally unrealistic (and with this size off cell
would require _even_ a couple of orders of more cores) - sorry but do you
know what you are doing? The pseudo potential would probably require much
higher a cut-off, did you test it? Do you plan to run magnetic systems -
nickel usually forms such.
I would recommend that you start from a much more trivial system and try
to understand these kind of calculations, and only then go for larger
systems and high-performance cluster.
Greetings from Paris,
Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
Ecole Normale Supérieure (ENS), Département de Chimie, Paris
Mobile (F) : +33 789 37 24 25 (CH) : +41 79 71 90 935
On Sat, 20 Jun 2015, Vishal Gupta wrote:
> Hi, I've been running an SCF calculation for a fee Ni system on High performance cluster but no matter how many processors I choose to run it, it always
> leads to segmentation fault :-
> mpirun noticed that process rank 0 with PID 6353 on node c7c exited on signal 11 (Segmentation fault).
> or excessive memory leakage.
> I mostly chose 6-20 processors but I've tried it once with 40 also.
> Input file :-
> calculation = 'scf' ,
> pseudo_dir = '/home/vishalgupta/GB' ,
> ibrav = 0,
> nat = 44,
> ntyp = 1,
> ecutwfc = 55 ,
> mixing_beta = 0.7 ,
> diagonalization = 'cg' ,
> CELL_PARAMETERS angstrom
> 22.271206224 0.000000000 0.000000000
> 0.000000000 11.150000000 0.000000000
> 0.000000000 0.000000000 1.500000000
> Ni 58.69340 Ni.pz-hgh.UPF
> ATOMIC_POSITIONS angstrom
> Ni 7.789220206286963 10.020142143601101 0.000000000000000
> Ni 0.000000000000000 0.000000000000000 0.000000000000000 ......
> K_POINTS automatic
> 40 20 9 0 0 0
> Can somebody tell me where the problem lies ?
> Thank you
> Vishal Gupta
> B.Tech. 2nd year Mechanical
> Indian Institute of Technology Ropar
> Rupnagar (140001), Punjab, India.
> Email :- vishal.gupta at iitrpr.ac.in
More information about the users