[Pw_forum] pw.x stuck

Duy Le ttduyle at gmail.com
Thu Jun 2 22:22:14 CEST 2011


Have you ever done vc-relax with a much smaller system? Say 1-3 atoms.
--------------------------------------------------
Duy Le
PhD Student
Department of Physics
University of Central Florida.

"Men don't need hand to do things"



On Thu, Jun 2, 2011 at 4:12 PM, Yongsheng Zhang <yshzhang88 at yahoo.com> wrote:
> Firstly, this system was running throughout using previous PWSCF version, v4.2.1.
>
> Further tests with v4.3 using different CPUs, pw.x is stuck at different places.
>
> (1) 1CUP: after one static iteration, it leaves error message,
>     WARNING: integrated charge=     0.00000000, expected=    64.00000000
>
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>     from electrons : error #         1
>     charge is wrong: smearing is needed
>  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
> (2) 4 CPUs: the system did relaxed. But when it run the "final scf", it is stuck at,
>    A final scf calculation at the relaxed structure.
>
>     The G-vectors are recalculated.
>
> (3) 8 CPUs: after one static iteration, it is stuck at,
>
>        k(    1) = (   0.1249057   0.1249057   0.1531359), wk =   0.2500000
>        k(    2) = (   0.1249057   0.3747172   0.1531359), wk =   0.5000000
>        k(    3) = (   0.3747172   0.3747172   0.1531359), wk =   0.2500000
>        k(    4) = (  -0.1249057   0.1249057  -0.1531359), wk =   0.2500000
>        k(    5) = (  -0.1249057   0.3747172  -0.1531359), wk =   0.5000000
>        k(    6) = (  -0.3747172   0.3747172  -0.1531359), wk =   0.2500000
>
> I am kind of pretty sure that there is something wrong in the compiling the code. Then, I did my best to change the make.sys file and recompile the code. The best results I can get so far is: pw.x do relax the system, but when it is trying to run "A final scf calculation at the relaxed structure", it stops at,
>     G cutoff = 2508.9230  ( 852531 G-vectors)     FFT grid: (108,108,180)
>     G cutoff = 1003.5692  ( 215507 G-vectors)  smooth grid: ( 64, 64,108)
>
>     Largest allocated arrays     est. size (Mb)     dimensions
>        Kohn-Sham Wavefunctions         1.65 Mb     (   3377,  32)
>        NL pseudopotentials             5.77 Mb     (   3377, 112)
>        Each V/rho on FFT grid          4.09 Mb     ( 268272)
>        Each G-vector array             0.81 Mb     ( 106570)
>        G-vector shells                 0.36 Mb     (  47252)
>     Largest temporary arrays     est. size (Mb)     dimensions
>        Auxiliary wavefunctions         6.60 Mb     (   3377, 128)
>        Each subspace H/S matrix        0.25 Mb     (    128, 128)
>        Each <psi_i|beta_j> matrix      0.05 Mb     (    112,  32)
>        Arrays for rho mixing          32.75 Mb     ( 268272,   8)
>
>
> And give the following error message:
> forrtl: severe (41): insufficient virtual memory
> Image              PC                Routine            Line        Source
> pw.x               0000000001124F0D  Unknown               Unknown  Unknown
> pw.x               0000000001123A15  Unknown               Unknown  Unknown
> pw.x               00000000010C2C39  Unknown               Unknown  Unknown
> pw.x               000000000105885D  Unknown               Unknown  Unknown
> pw.x               000000000109736E  Unknown               Unknown  Unknown
> pw.x               000000000053EEBF  allocate_nlpot_            68  allocate_nlpot.f90
> pw.x               0000000000454D76  init_run_                  60  init_run.f90
> pw.x               00000000004716B7  move_ions_                290  move_ions.f90
> pw.x               000000000044AA55  MAIN__                    173  pwscf.f90
> pw.x               000000000044A60C  Unknown               Unknown  Unknown
> libc.so.6          00002AE8FFD68974  Unknown               Unknown  Unknown
> pw.x               000000000044A519  Unknown               Unknown  Unknown
>
>
> I am wondering that why the "insufficient virtual memory" happens at the last step. If there is no sufficient virtual memory, the code should not even run at the first iteration. While, I collect the relaxed structure and re-run pw.x, it works without any problem.
>
> Thanks
> Yongsheng
>
>
> --- On Thu, 6/2/11, Paolo Giannozzi <giannozz at democritos.it> wrote:
>
>> From: Paolo Giannozzi <giannozz at democritos.it>
>> Subject: Re: [Pw_forum] pw.x stuck
>> To: "PWSCF Forum" <pw_forum at pwscf.org>
>> Date: Thursday, June 2, 2011, 1:45 AM
>>
>> On Jun 1, 2011, at 19:05 , Jack London wrote:
>>
>> > Yes, I have tried many times
>>
>> many times in the same conditions is better than nothing,
>> but you should try in different conditions as well.
>>
>> P.
>> ---
>> Paolo Giannozzi, Dept of
>> Chemistry&Physics&Environment,
>> Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
>> Phone +39-0432-558216, fax +39-0432-558222
>>
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


More information about the Pw_forum mailing list