[Pw_forum] Ph.x code help

Vincenzo Verdolino vincenzo.verdolino at gmail.com
Wed Aug 27 18:31:00 CEST 2014


I don't know for sure but and i encourage someone else to give a better explenation if any. However, i suggest two things:

1) in the scf you can use as many nodes and processors you need. You might want to use
wf_collect=.true.
It works in writing and not in the reading process and it avoids conflicts between pw.x and ph.x
2) i personally do not use the wavefunction collector because once i find the best number of nodes and processors in terms of scalability (but this is architecture dependent) i freez it forever for that system. So you might want to try and error apprach i.e. 1 proc- 10 proc-20proc but it debends on the architecture of your nodes. Still i suggest to use 1 node and n processors which you are going to define by testing.
3) you found that problem because of the very small number of k-points. Just to give e try use

K_POINTS automatic
4 4 4 0 0 0

In the scf run and try to use more processors in your ph.x

Sent from my iPhone

> On 27/ago/2014, at 15:44, "Samin, Adib J." <samin.2 at buckeyemail.osu.edu> wrote:
> 
> Actually, I was running it with two processors on one node but after trying it with one processor it worked. So thanks again for all your help.
> 
> In any case, I was just doing this example to get familiar with the procedure and I am planning to use it for my own research. In that case the simulation will be more complicated involving many atoms. How would I know how many processors to use?
> 
> Is there a general rule of thumb that I can use to help guide me for the future?
> 
> 
> 
> Thanks,
> 
> Adib Samin
> 
> 
>  
> 
>  
> 
> 
>  
> From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of Vincenzo Verdolino <vincenzo.verdolino at gmail.com>
> Sent: Wednesday, August 27, 2014 3:53 AM
> To: PWSCF Forum
> Subject: Re: [Pw_forum] Ph.x code help
>  
> did you try to run both the scf and the ph.x job with 1 node and 1 proc?
> 
> 
>> On Tue, Aug 26, 2014 at 6:32 PM, Samin, Adib J. <samin.2 at buckeyemail.osu.edu> wrote:
>> Yes it is.
>> 
>> Thanks,
>> 
>> Adib Samin
>> 
>> 
>> 
>> 
>> 
>> 
>> ________________________________________
>> From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of Vincenzo Verdolino <vincenzo.verdolino at gmail.com>
>> Sent: Tuesday, August 26, 2014 12:12 PM
>> To: PWSCF Forum
>> Subject: Re: [Pw_forum] Ph.x code help
>> 
>> Quick question
>> 
>> Is the Pseudo Potential the one suggested in the test input file? I see you are usin a pz one...i don't know but could you double check that??
>> 
>> Sent from my iPhone
>> 
>> > On 26/ago/2014, at 17:53, "Samin, Adib J." <samin.2 at buckeyemail.osu.edu> wrote:
>> >
>> > Dear Quantum Espresso Users,
>> >
>> > Thanks again for all your help. I was finally able to run the ph.x code but I am still not generating the dynamical matrix but rather a blank file.
>> > I have attached the output of the ph.x code for the input files below. Any help on this matter would be appraciated.
>> >
>> > Atomic displacements:
>> >     There are   2 irreducible representations
>> >
>> >     Representation     1      3 modes -T_2g G_25' G_5+  To be done
>> >
>> >     Representation     2      3 modes -T_1u G_15  G_4-  Not done in this run
>> >
>> >     Compute atoms:     1,    2,
>> >
>> >
>> >
>> >     Alpha used in Ewald sum =   0.7000
>> >     PHONON       :     0.56s CPU         0.61s WALL
>> >
>> >
>> >
>> >     Representation #  1 modes #   1  2  3
>> >
>> >     Self-consistent Calculation
>> >
>> >      iter #   1 total cpu time :     0.9 secs   av.it.:   5.0
>> >      thresh= 0.100E-01 alpha_mix =  0.700 |ddv_scf|^2 =  0.189E-06
>> >
>> >      iter #   2 total cpu time :     1.3 secs   av.it.:   9.5
>> >      thresh= 0.434E-04 alpha_mix =  0.700 |ddv_scf|^2 =  0.320E-08
>> >
>> >      iter #   3 total cpu time :     1.7 secs   av.it.:   9.5
>> >      thresh= 0.566E-05 alpha_mix =  0.700 |ddv_scf|^2 =  0.278E-10
>> >
>> >      iter #   4 total cpu time :     2.1 secs   av.it.:   8.8
>> >      thresh= 0.528E-06 alpha_mix =  0.700 |ddv_scf|^2 =  0.296E-13
>> >
>> >      iter #   5 total cpu time :     2.5 secs   av.it.:   9.4
>> >      thresh= 0.172E-07 alpha_mix =  0.700 |ddv_scf|^2 =  0.287E-15
>> >
>> >     End of self-consistent calculation
>> >
>> >     Convergence has been achieved
>> >
>> >     Stopping because representation    2 is not done
>> >
>> >
>> > Thanks,
>> >
>> > Adib Samin
>> >
>> >
>> > Adib Samin
>> > Postdoc
>> > The Department of Aerospace and Mechanical Engineering
>> > The Ohio State University
>> >
>> >
>> >
>> > ________________________________________
>> > From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of Samin, Adib J. <samin.2 at buckeyemail.osu.edu>
>> > Sent: Tuesday, August 26, 2014 11:08 AM
>> > To: PWSCF Forum
>> > Subject: Re: [Pw_forum] Ph.x code help
>> >
>> > Thank you so much for all your help. I was finally able to run the code by reducing the number of processors being used.
>> > I still have one question however. The si.dynG file I am creating seems to be a blank file. Is that correct?
>> > I am currently attempting to run dynmat.x to diagonalize the matrix but I am concerned because I don't see the matrix.
>> >
>> > Thanks,
>> >
>> > Adib Samin
>> >
>> >
>> >
>> >
>> >
>> >
>> > ________________________________________
>> > From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of Vincenzo Verdolino <vincenzo.verdolino at gmail.com>
>> > Sent: Tuesday, August 26, 2014 10:53 AM
>> > To: PWSCF Forum
>> > Subject: Re: [Pw_forum] Ph.x code help
>> >
>> > 12 pros is fine for the scf but try to change it for ph.x
>> >
>> > Sent from my iPhone
>> >
>> >> On 26/ago/2014, at 16:19, "Samin, Adib J." <samin.2 at buckeyemail.osu.edu> wrote:
>> >>
>> >> Thank you for your response. I am still getting errors running the code. I am indeed running it in parallel using 12 processors. How do I check for the consistencies mentioned below?
>> >> Any clarification would be greatly appreciated.
>> >>
>> >> Thanks,
>> >>
>> >> Adib Samin
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >> ________________________________________
>> >> From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of Ari P Seitsonen <Ari.P.Seitsonen at iki.fi>
>> >> Sent: Tuesday, August 26, 2014 9:59 AM
>> >> To: PWSCF Forum
>> >> Subject: Re: [Pw_forum] Ph.x code help
>> >>
>> >> Dear Adib Samin,
>> >>
>> >>  How did you try to run the example, parallellising over the
>> >> representations/q vectors? If yes, how many MPI tasks did you ask for, how
>> >> many partitions for the representations/q vectors? Please check whether
>> >> these are consistent.
>> >>
>> >>    Greetings from Sunny Montrouge,
>> >>
>> >>       apsi
>> >>
>> >> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
>> >>  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 Tue, 26 Aug 2014, Samin, Adib J. wrote:
>> >>>
>> >>>
>> >>> Thank you for your help.
>> >>>
>> >>> I have attached below the scf.in file. I hope that helps in diagnosing the
>> >>> problem.
>> >>>
>> >>> &control
>> >>>   calculation='scf'
>> >>>   restart_mode='from_scratch',
>> >>>   pseudo_dir='./'
>> >>>   outdir='/nfs/11/osu7834/'
>> >>>   prefix='SIPH',
>> >>> /
>> >>> &system
>> >>>   ibrav=2, celldm(1)=10.20, nat=2, ntyp=1,
>> >>>   ecutwfc =16.0
>> >>> /
>> >>> &electrons
>> >>>   conv_thr =  1.0d-9
>> >>>   mixing_beta = 0.7
>> >>> /
>> >>> ATOMIC_SPECIES
>> >>> Si  28.0855 Si.pz-vbc.UPF
>> >>> ATOMIC_POSITIONS
>> >>> Si 0.00 0.00 0.00
>> >>> Si 0.25 0.25 0.25
>> >>> K_POINTS
>> >>>  10
>> >>>   0.1250000  0.1250000  0.1250000   1.00
>> >>>   0.1250000  0.1250000  0.3750000   3.00
>> >>>   0.1250000  0.1250000  0.6250000   3.00
>> >>>   0.1250000  0.1250000  0.8750000   3.00
>> >>>   0.1250000  0.3750000  0.3750000   3.00
>> >>>   0.1250000  0.3750000  0.6250000   6.00
>> >>>   0.1250000  0.3750000  0.8750000   6.00
>> >>>   0.1250000  0.6250000  0.6250000   3.00
>> >>>   0.3750000  0.3750000  0.3750000   1.00
>> >>>   0.3750000  0.3750000  0.6250000   3.00
>> >>>
>> >>>
>> >>> Thanks,
>> >>>
>> >>> Adib Samin
>> >>>
>> >>>
>> >>> Adib Samin
>> >>>
>> >>> Postdoc
>> >>>
>> >>> The Department of Aerospace and Mechanical Engineering
>> >>>
>> >>> The Ohio State University
>> >>>
>> >>>
>> >>>
>> >>>
>> >>> ​
>> >>>
>> >>>
>> >>>
>> >>> ____________________________________________________________________________
>> >>> From: pw_forum-bounces at pwscf.org <pw_forum-bounces at pwscf.org> on behalf of
>> >>> Vincenzo Verdolino <vincenzo.verdolino at gmail.com>
>> >>> Sent: Tuesday, August 26, 2014 5:38 AM
>> >>> To: PWSCF Forum
>> >>> Subject: Re: [Pw_forum] Ph.x code help
>> >>> Dear,
>> >>>
>> >>> just make sure you
>> >>>
>> >>> cp -r -f SIPH.save /nfs/11/osu7834/
>> >>>
>> >>> sometimes it happens that your outir in the scf calculation is not the same
>> >>> specified in your ph.x calculation
>> >>>
>> >>> could you also paste the scf.in job file?
>> >>>
>> >>> thanks
>> >>>
>> >>> vincenzo
>> >>>
>> >>>
>> >>> On Mon, Aug 25, 2014 at 10:07 PM, Samin, Adib J.
>> >>> <samin.2 at buckeyemail.osu.edu> wrote:
>> >>>
>> >>>     Dear Quantum Espresso users,
>> >>>
>> >>>
>> >>>     I am trying to run the ph.x executable using a simple silicon
>> >>>     example which I found online. I first perform the scf
>> >>>     calculation and save the results and then use the same
>> >>>     outdirectory and prefix when running the ph.x code. However, I
>> >>>     keep getting the same error which says:
>> >>>
>> >>>     'Error in routine image_q_irr (1):
>> >>>
>> >>>       some images have no rapp'
>> >>>
>> >>>     I have attached the ph input file.
>> >>>
>> >>>     I would really appreciate it if someone can help explain to me
>> >>>     why this is happening.
>> >>>
>> >>>
>> >>> Phonons of Si at Gamma
>> >>> &inputph
>> >>> tr2_ph=1.0d-14,
>> >>> amass(1)=28.0855,
>> >>> prefix='SIPH',
>> >>> outdir='/nfs/11/osu7834/'
>> >>> fildyn='si.dynG',
>> >>> /
>> >>> 0.0 0.0 0.0
>> >>>
>> >>>
>> >>>
>> >>>     Thanks,
>> >>>
>> >>>     Adib Samin
>> >>>
>> >>>
>> >>>
>> >>>
>> >>>     Adib Samin
>> >>>
>> >>>     Postdoc
>> >>>
>> >>>     The Department of Aerospace and Mechanical Engineering
>> >>>
>> >>>     The Ohio State University
>> >>>
>> >>>
>> >>> _______________________________________________
>> >>> Pw_forum mailing list
>> >>> Pw_forum at pwscf.org
>> >>> http://pwscf.org/mailman/listinfo/pw_forum
>> >>
>> >> _______________________________________________
>> >> Pw_forum mailing list
>> >> Pw_forum at pwscf.org
>> >> http://pwscf.org/mailman/listinfo/pw_forum
>> >
>> > _______________________________________________
>> > Pw_forum mailing list
>> > Pw_forum at pwscf.org
>> > http://pwscf.org/mailman/listinfo/pw_forum
>> >
>> > _______________________________________________
>> > Pw_forum mailing list
>> > Pw_forum at pwscf.org
>> > http://pwscf.org/mailman/listinfo/pw_forum
>> >
>> > _______________________________________________
>> > Pw_forum mailing list
>> > Pw_forum at pwscf.org
>> > http://pwscf.org/mailman/listinfo/pw_forum
>> 
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://pwscf.org/mailman/listinfo/pw_forum
>> 
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://pwscf.org/mailman/listinfo/pw_forum
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20140827/75986553/attachment.html>


More information about the users mailing list