Skip to Content.
Sympa Menu

forum - Re: [abinit-forum] How to reduce number of G vectors in GW?

forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] How to reduce number of G vectors in GW?


Chronological Thread 
  • From: Anglade Pierre-Matthieu <anglade@gmail.com>
  • To: forum@abinit.org
  • Subject: Re: [abinit-forum] How to reduce number of G vectors in GW?
  • Date: Fri, 6 Mar 2009 09:26:12 +0100
  • 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:content-transfer-encoding; b=pEGERPDslprO/0U/Sk0tFEbmAavK0QRYe5zvZPCq8/RTb5dz6RPal01kPuaCDVqUqv 8W9zbVGTC/TQYJ5WTqryuNpCcS0Y3FA80Ry0Fh36FpQnxBywarBnnBD5dABXxCiAI1gQ V3+wvfeLtZ7BnEMb2TiKQ+uXHqH779noLLQKc=

Unfortunately, the number of G vectors depends on ecut and the cell size ...

On Thu, Mar 5, 2009 at 8:07 PM, Aaron Zhou <fishzhou@gmail.com> wrote:
> The ecut I was using is 12 hartree, which is already very small...you
> mean that the number of g vectors ONLY depend on ecut?
>
> On Thu, Mar 5, 2009 at 2:01 PM, lan haiping <lanhaiping@gmail.com> wrote:
>>
>> Do you ever check your gs results ?
>> I think the number of G  depends on  the number of planewaves you used for
>> calculation. Therefore, it should depend on your ecut for GS calculation.
>> - Show quoted text -
>> On Fri, Mar 6, 2009 at 2:15 AM, <fishzhou@gmail.com> wrote:
>>>
>>> Dear all,
>>>
>>> When I was doing a GW calculation on a big super cell, the number of G
>>> vectors
>>> is too big that I cannot handle.
>>>
>>> I found that by reducing ngkpt, toldfe, nband or nbandkss did not help too
>>> much. Does anyone know that how I can reduce the accuracy, I mean reduce
>>> the
>>> number of G vectors so that I can lower my computational time.
>>>
>>> Thanks a lot
>>>
>>> Aaron
>>>
>>> P.S.
>>> Below are the part of the input file and output
>>>
>>> =======================================
>>> ndtset 2
>>> jdtset 1 2
>>>
>>> # Dataset 1 : usual self-consistent calculation for ground state charge
>>> density
>>> kptopt1   1
>>> ngkpt1    1 1 4
>>> nshiftk1  1
>>> shiftk1   0.0 0.0 0.0
>>>
>>> toldfe1  1.0d-6    # Will stop when this tolerance is achieved on total
>>> energy
>>> prtden1  1
>>>
>>> # Dataset 2: calculation of kss file
>>> kptopt2   1
>>> ngkpt2    1 1 4
>>> nshiftk2  1
>>> shiftk2   0.0 0.0 0.0    # This grid contains the Gamma point
>>> istwfk2   5*1
>>>
>>> iscf2     -2             # Non self-consistent calculation
>>> getden2   -1             # Read previous density file
>>> tolwfr2   1.0d-9         # Still get it converged
>>> nband2    160
>>>
>>> nbandkss2 192        # Number of bands in KSS file (the maximum possible)
>>>
>>>
>>> ======================================================
>>>            ****                  26480              12.918
>>>            ****                  26481              12.918
>>>            ****                  26482              12.918
>>>            ****                  26483              12.967
>>>            ****                  26484              12.967
>>>            ****                  26485              12.967
>>>            ****                  26486              12.967
>>>
>>>
>>>  Since the number of g's to be written on file
>>>  was 0 or too large, it has been set to the max. value.,
>>>  computed from the union of the sets of G vectors for the different
>>> k-points.
>>>  Number of g-vectors written on file is: 26486
>>>  Number of bands written on file is:   192
>>>
>>>  Since the number of bands to be computed
>>>  is less than the number of G-vectors found,
>>>  the program will perform partial diagonalizations.
>>>
>>>  Amount of disk space required by _STA file=  233.50 Mbytes.
>>>
>>>  Opening file for KS structure output: gwo_DS2_KSS
>>>  number of Gamma centered plane waves  26486
>>>  number of Gamma centered shells  26486
>>>  number of bands    192
>>>  maximum angular momentum components      3
>>>  number of symmetry operations  1 (without inversion)
>>>   *   g( 1)=  0  0  0
>>>   *   g( 2)=  1  0  0
>>>   *   g( 3)= -1  0  0
>>>   *   g( 4)=  0  1  0
>>>   *   g( 5)=  0 -1  0
>>>   *   g( 6)=  1  1  0
>>>   *   g( 7)= -1  1  0
>>>   *   g( 8)=  1 -1  0
>>>   *   g(****)= -1  1-13
>>>   *   g(****)=  1 -1-13
>>>   *   g(****)= -1 -1-13
>>>   *   g(****)=  2  0-13
>>>   *   g(****)= -2  0-13
>>>   *   g(****)=  0  2-13
>>>   *   g(****)=  0 -2-13
>>> -P-0000
>>> -P-0000  k-point   1
>>> -P-0000  Calculating <G|H|G'> elements
>>> -P-0000
>>> -P-0000  Begin partial diago for ikpt=  1              - Size of
>>> mat.=25725 - #
>>> bnds=  192
>>
>>
>>
>> --
>> Hai-Ping Lan
>> Department of Electronics ,
>> Peking University , Bejing, 100871
>> lanhaiping@gmail.com, hplan@pku.edu.cn
>>
>
>



--
Pierre-Matthieu Anglade




Archive powered by MHonArc 2.6.15.

Top of Page