Skip to Content.
Sympa Menu

forum - Re: [abinit-forum] memory allocation issues

forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] memory allocation issues


Chronological Thread 
  • From: Mohua Bhattacharya <mohua.simoom@gmail.com>
  • To: forum@abinit.org
  • Subject: Re: [abinit-forum] memory allocation issues
  • Date: Fri, 19 Feb 2010 21:37:26 -0500
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=IRfRnFm5NbRRROVbnj2N9K7uvlbFlRqsnL0OyLUQDFri69Z6W1aNOSjIzKEbTPqYMz qyYfTZseEd7QxgJ3tf6PLLlmh1izKplHzDKX+A+/3vtVnIMouTeECLWj1YVcr+uCASKX VDe0dx7EWHDmQ+OOn/1XOnfi/lTfkvR2N7Iko=

Dear Matteo,

Thank you very much for your reply. I have another question which is kind  of trivial.  Do I need to change to a 64 bit version as well?

Thanks again,
Mohua

On Fri, Feb 19, 2010 at 3:41 PM, Matteo Giantomassi <Matteo.Giantomassi@uclouvain.be> wrote:
> Hello,
>
> I am trying to do a structural relaxation of a 2*2*2 KNbO3 supercell . The
> problem is that it doesn't find enough memory and it exits. I was
> wondering
> if  there  is a way to get around this problem ? Waiting for your inputs.
>
> Thanks for your time,
>
> Mohua
>
>  snippets of the output file:


 Dear Mohua,

 The standard k-point parallelism is of no help to reduce
 the memory requirements since your system has only one k-point,
 You might try the band-fft parallelism.
 Please have a look at the paral_kgb input variable described here

http://www.abinit.org/documentation/helpfiles/for-v5.8/input_variables/varpar.html#paral_kgb

 Best Regards
 Matteo Giantomassi



>
> memory : analysis of memory needs
> ================================================================================
>  Values of the parameters that define the memory need of the present run
>    intxc =         0  ionmov =         2    iscf =         7     ixc =
>  11
>   lmnmax =         4   lnmax =         4   mband =       330  mffmem =
>   1
> P  mgfft =       216   mkmem =         1 mpssoang=         4     mpw =
>  392872
>   mqgrid =      3001   natom =        80    nfft =   6718464    nkpt =
>   1
>   nloalg =         4  nspden =         1 nspinor =         1  nsppol =
>   1
>     nsym =        64  n1xccc =         0  ntypat =         3  occopt =
>   1
> ================================================================================
> P This job should need less than                    4141.509 Mbytes of
> memory.
>   Rough estimation (10% accuracy) of disk space for files :
>   WF disk file :   1978.270 Mbytes ; DEN or POT disk file :     51.260
> Mbytes.
> ================================================================================
>
>  Biggest array : cg(disk), with   1978.2698 MBytes.
> -P-0000  leave_test : synchronization done...
>  memana : allocated an array of   1978.270 Mbytes, for testing purposes.
> -P-0000
> -P-0000  memana : ERROR -
> -P-0000   Test failed to allocate    4141.509 Mbytes
> -P-0000   It is not worth to continue
> -P-0000   Action : modify input variable to fit the available memory.
> -P-0000   or increase limit on available memory.
> -P-0000
> -P-0000  leave_new : decision taken to exit ...
> -P-0000  leave_new : synchronization done...
> -P-0000  leave_new : exiting...
> abinip(78007) malloc: *** mmap(size=542900224) failed (error code=12)
> *** error: can't allocate region
> *** set a breakpoint in malloc_error_break to debug
> abinip(78007) malloc: *** mmap(size=542900224) failed (error code=12)
> *** error: can't allocate region
> *** set a breakpoint in malloc_error_break to debug
> abinip(78007) malloc: *** mmap(size=542900224) failed (error code=12)
> *** error: can't allocate region
>






Archive powered by MHonArc 2.6.16.

Top of Page