[Pw_forum] DOS and pDOS parallelization

[Madhura Marathe]

 Thanks for the explanation. That was really helpful.

 -Madhura

>
> Dear Madhura,
>
>>
>>  Thanks for the explanation.
>>  I had always thought that DOS code also requires wavefunction files;
>> and
>> that is why one needs to give the same prefix and outdir path as in scf
>> calculations. If code does not need wavefunctions, which files it uses
>> for DOS calc.s?
>
> the code needs mainly the eigenvalues, which are stored in
> $prefix.save/K?/eigenval.xml  files,
> where K? (i.e. K1, K2, etc ) correspond to each kpt.
>
>
>>  As per suggestion, I have run PDOS calc.s with the same number of pools
>> and processors (in case of B) and got results. That means PDOS code is
>> parallelized with pool implementation, but not DOS.
>>
>
> yes you are right,
> this means that pDOS is parallelized over G vectors and pools, while
> DOS calculation is not parallelized but is anyway compatible with
> the parallel environment (i.e. it does not crash if run parallel)
>
> andrea
>
>>  Thanks once again,
>>  Madhura.
>>
>> >
>> > Dear Madhura,
>> >
>> > the issue of using the same parallelism scheme (# of processors, # of
>> > pools, etc) in postprocessing calculations is mainly concerned with
>> the
>> > need to read wfcs. As you wrote, if you specify wf_collect = .TRUE.
>> you
>> > completely remove this issue (you can read wfcs whatever the
>> > parallelization
>> > scheme you used to produce them).
>> >
>> >>From this point of view, the main difference between DOS and pDOS
>> > calculations is that the former does not need to read wfcs, while the
>> > latter does.
>> >
>> > Moreover, the taks the DOS program has to perform is so inexpensive
>> that
>> > it is performed by a single processor even if the code is run in
>> parallel.
>> > This is consistent with your observation at (ii).
>> >
>> > In the case of pDOS (i.e. when running projwfc.x) if you do not
>> collect
>> > wfcs you should run with the same # of processors and the same pool
>> > parallelism. Otherwise, as you experienced, you get a davcio error
>> > (some of the wfcs files were not found).
>> > According to me, anyway, projwfc should work with pools if the above
>> > conditions are fulfilled.
>> >
>> > hope it helps
>> > andrea
>> >
>> >
>> >>
>> >>  There have been many discussions in this forum about parallelization
>> of
>> >> DOS and projected DOS codes. However, some of the points were not
>> clear
>> >> to
>> >> me, so I have performed some scf calculations for a very simple
>> system
>> >> using 8 processors and with flag wf_collect 'false'. I have performed
>> >> same
>> >> calculations once without use of pools (A) and then using 2 pools
>> (B).
>> >> Then using theses wavefunctions I have performed DOS and PDOS
>> >> calculations. Following is the summary of results and my
>> interpretation:
>> >>
>> >> i) Pool parallelization is not implemented for both these codes.
>> >>
>> >> ii) DOS calculations: for both the cases A and B, one can calculate
>> DOS
>> >> with the same no. of processors (= 8) and the results match within
>> >> numerical errors; even though for case B, the wavefunctions were
>> >> obtained
>> >> with pooling and DOS without it.
>> >>  For case A, even if I use 4 processors I get identical results as
>> when
>> >> I
>> >> use 8 processors. (Note, I have not checked with less no. of proc.s
>> for
>> >> B).
>> >> => The condition that we need the same no. of processors and pools as
>> >> were
>> >> used in scf calculations is not necessary; and it is possible to get
>> DOS
>> >> results even with wavefunctions generated with pool parallelization.
>> >>
>> >> iii) PDOS calculations: This can be calculated only in case of A and
>> >> using
>> >> the same no. of processors. If I use wavefunctions generated in case
>> B
>> >> or
>> >> less no. of processors (= 4) with A wavefn.s then I get "davcio"
>> error.
>> >> =>
>> >> For PDOS calculations, one cannot use wavefunctions generated with
>> pool
>> >> parallelization unless wf_collect flag is set to 'true' for scf
>> calc.s;
>> >> also one has to use the same no. of proc.s as were used for scf
>> calc.s
>> >> to
>> >> get projected DOS.
>> >>
>> >> Now my question is are these interpretations correct? Or they may
>> change
>> >> for some other system?? Do I need to do some more checks to ascertain
>> >> these? If yes, what sort of tests?
>> >>
>> >> Thanks for reading this long mail patiently, but I need to clarify on
>> >> these points before I can start with bigger systems.
>> >> Sincerely,
>> >> Madhura.
>> >>
>> >>
>> >>
>> >
>> > --
>> > Andrea Ferretti
>> > MIT, Dept Material Science & Engineering
>> > bldg 13-4078, 77, Massachusetts Ave, Cambridge, MA
>> > Tel: +1 617-452-2455;  Skype: andrea_ferretti
>> > URL: http://quasiamore.mit.edu
>> >
>> > Please, if possible, don't send me MS Word or PowerPoint attachments
>> > Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html
>> >
>> >
>> > ___
>> > 

[Pw_forum] compile error, pathscale compiler

[Paolo Giannozzi]

On Thursday 13 August 2009 20:18, Jess Kondor wrote:

> CALL synchronize( gid )
>  ^
> pathf95-700 pathf90: ERROR BCAST_REAL, File = mp_base.f90, Line = 62,
> Column = 14
>   The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

apparently some genius working at pathscale has decided that "synchronize" 
is a nice name for an intrinsic function. Define it external or change the
routine name.

P.
-- 
Paolo Giannozzi, Democritos and Udine University

[Pw_forum] negative (alpha)^2*F(w)

[Stefano Baroni]

As discussed, many times in this forum, as well as in many textbooks,  
negative squared ferquencies are a signal of a mechanical instability.  
Are you sure that you want to spend a lot of computer resources  
calculating the supercondictive properties of an instable (and, hence,  
non-existent unless very special conditions occur) system? SB

---
swift text written and sent on the go

On 13/ago/2009, at 11.00, Ali Tavana  wrote:

> Dear QE users,
>
> I have a question about e-ph lambda, calculated by "matdyn" program.  
> I've found that for some frequencies the quantity "(alpha)^2*F(w)/w"  
> or "dos_a2F(w)/w" can be negative, most probably because of the  
> negative (imaginary) frequencies (I can't believe a2F(w) become  
> negative!). In "matdyn" program these negative values are considered  
> in lambda calculations. Should one remove these negative values as  
> these imaginary frequency modes can never exist in the system? then  
> where are those phonon branches? I think negative values of  
> frequency can show in some manner more strong electron structure  
> interaction, so what do you suggest one should do?
>
> regards,
> Ali
>
>
>
>
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

[Pw_forum] Trying to clarify the dipole correction (again!)

[J. J. Ramsey]

- Original Message 

> From: Stefano de Gironcoli 
> To: PWSCF Forum 
> Sent: Thursday, August 13, 2009 11:59:14 AM
> Subject: Re: [Pw_forum] Trying to clarify the dipole correction (again!)
> 
> Dear J.J. Ramsey
> 
>  at variance with what is defined in Bengtsson's paper the dipole 
> potential is not discontinuous  in QE but is defined by periodically 
> alternating an increasing slope (associate to the "physical" electric 
> field present in the cell) and a  rapidly decreasing slope that should 
> be located in a region of negligibly small charge density. This is done 
> to avoid too sharp features in the calculation.

In other words, the discontinuity in Bengtsson's paper, where the slope is 
infinite, is replaced by a large but finite slope.

>so if emaxpos = 0.45 and eopreg=0.1 the region where the correction 
> is physically meaningful goes from -45% to +45% of the cell and the 
> regression  region  (that should contains no charge density)  occupies 
> the central 10% of the cell. This is a good setting if you center your 
> slab around the origin.

I gather you mean the case where the slab is near the bottom of the unit cell.



  

[Pw_forum] Trying to clarify the dipole correction (again!)

[Stefano de Gironcoli]

Dear J.J. Ramsey

 at variance with what is defined in Bengtsson's paper the dipole 
potential is not discontinuous  in QE but is defined by periodically 
alternating an increasing slope (associate to the "physical" electric 
field present in the cell) and a  rapidly decreasing slope that should 
be located in a region of negligibly small charge density. This is done 
to avoid too sharp features in the calculation. The definition is also 
generalized so that the region where the dipole correction operates can 
be centered in an arbitrary point of the cell and not just in the center 
of it as defined in the paper.
the regression region (the region with a rapidly decreasing slope) 
occupies a fraction eopreg of the cell size while the maximum of the 
increasing slope occurs at a fraction emaxpos.
   so if emaxpos = 0.45 and eopreg=0.1 the region where the correction 
is physically meaningful goes from -45% to +45% of the cell and the 
regression  region  (that should contains no charge density)  occupies 
the central 10% of the cell. This is a good setting if you center your 
slab around the origin.
   if instead emaxpos=0.95 and eopreg=0.1 the the dipole correction is 
applied for +5% to +95% of the cell while in the rest one have the
regression region. this is OK if you put your slab in the center of the 
cell.

  Stefano de Gironcoli - SISSA and DEMOCRITOS


J. J. Ramsey wrote:
> In the reference on the dipole correction used in QE, L. Bengtsson, PRB 59, 
> 12301 (1999), the dipole potential is given as
>
> V_dip(z) = 4*pi*m*(z/z_m - 0.5), 0 < z < z_m
>
> where z_m is the height of the box, and there is a discontinuity in the 
> potential at z = 0,z_m. I'm not sure how this fits into the QE documentation 
> on the dipole correction. Is V_dip in QE supposed to be something like this?:
>
> V_dip(z) = 4*pi*m*[z/z_m - (emaxpos + eopreg -1)], emaxpos + eopreg -1 < 
> z/z_m < emaxpos;
>  = something with a steep but finite slope, otherwise
>
> My guess is probably wrong, but the explanations in the documentation of the 
> roles of emaxpos and eopreg appear to be tied to the case where eamp  is 
> nonzero. I'm not sure if the documentation for the dipfield variable is 
> supposed to indicate that the actual slope of the potential is supposed to be 
> eamp + 4*pi*m/z_m.
>
> University of Akron
> Civil Engineering (!) Dept.
>
>
>   
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>   

[Pw_forum] Frequency and zero point energy calculations

[oulihui666]

ega(  8 -  8) =   1299.1  [cm-1]   --> A_1g X_1  M_1   R
 omega(  9 -  9) =   2335.0  [cm-1]   --> A_2u X_4' M_4'  I
 **
 PHONON   : 1h48m CPU time,2h25m wall time
..
There are still negative frequency in output file by increasing K_POINTS and 
reducing tr2_ph, anyone who help me will be appreciated. 
In addition, I intend to calculate the zero point energy of system by PHNONS 
code, but it does not give the zero point energy and other thermodynamics 
properties in output file. How could I do? I need your help urgently.
 
Thanks in advance 
Best regards 
Lihui 

--

==
Lihui Ou
PH.D Candidate in Electrochemistry  
College of Chemistry and Molecular Science
Wuhan University,430072,Hubei Province,China 
E-mail:oulihui666 at 126.com
==
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/265ef2f1/attachment-0001.htm
 

[Pw_forum] Frequency and zero point energy calculations

[Eyvaz Isaev]

Dear Lihui,

>?performing calulation of frequency and zero point
> energy of CO2 molecule. However, unfortunately, I have
> obtained?negative frequency.

> there are several possible reseaons: 

Well, you did not include one more reason counting them: 
there is a physical reason for this as you consider  CO2 molecule (the linear 
chain). Obviously, its physical properties do not change if you rotate it 
fixing C atom in the cell center. This might lead to negative frequencies. 

If you like to consider CO2 as a molecule you should use the Gamma point only 
(i.e. you consider CO2 in a big box).

> In addition, I intend to calculate the zero point
> energy of system by PHNONS code, but it does not give the
> zero point energy and other thermodynamics properties in
> output file. How could I do? I need your help
> urgently.

PH code itself does not calculate these properties. 
You can try download the QHA code from qe-forge.org. If some troubles with 
this, please let me know.

>  I need your help urgently.
Please keep in mind our forum consists of volunteers and nobody is oblidged to 
give you urgent reply.


Bests,
Eyvaz. 
P.S. Can you please show eigendisplacements? 

---
Prof. Eyvaz Isaev, 
Theoretical Physics Department, Moscow State Institute of Steel & Alloys, 
Russia, 
Department of Physics, Chemistry, and Biology (IFM), Linkoping University, 
Sweden 
Condensed Matter Theory Group, Uppsala University, Sweden 
Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com



  

[Pw_forum] scf

[spmmal]

?lvaro Alves wrote:
> Hello I am doing a scf calculation system and uses degauss = 0.003 and 
> needed change it paar 0001. With this new value, the calculation does 
> not converge. What may be happening?
>
> Regards, Alvaro
>
>
>
> 
> Veja quais s?o os assuntos do momento no Yahoo! + Buscados: Top 10 
> 
>  
> - Celebridades 
> 
>  
> - M?sica 
> 
>  
> - Esportes 
> 
>  
>
> 
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>   
Is there any general rule of setting parameters to make the SCF 
calculation  converge. Thanks.

[Pw_forum] running the parallel code

[Paolo Giannozzi]

On Saturday 08 August 2009 19:40, vamsi kv wrote:

> I want to know how to run pwscf on parallel systems with shared memory.
> what type of libraries has to be used.

use MPI 
-- 
Paolo Giannozzi, Democritos and Udine University

[Pw_forum] difference in vc-relax and relax calculations

[Bipul Rakshit]

Dear PWSCF users,
I
I am doing calculation of LaMnO3. First I take the experimental lattice
constant and atomic positions, and then run first vc-relax. After the system
is relaxed. I use those lattice parameters and atomic position to do a relax
calculation.

But after running the vc-relax calculation i got appreciable very less
pressure(~ -0.69 kbar), but as soon as I run the relax calculation, I am
getting pressure of about -152.0 kbar. I didnt understand the error. Please
have a look at my files and suggest me.
regards
-- 
Dr. Bipul Rakshit
Research Associate,
S N Bose Centre for Basic Sciences,
Salt Lake,
Kolkata 700 098
India
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/830f52ea/attachment.htm
 
-- next part --
A non-text attachment was scrubbed...
Name: lamno3.vc-relax.in
Type: application/octet-stream
Size: 3809 bytes
Desc: not available
Url : 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/830f52ea/attachment.obj
 
-- next part --
A non-text attachment was scrubbed...
Name: lamno3.vc-relax.out
Type: application/octet-stream
Size: 2323 bytes
Desc: not available
Url : 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/830f52ea/attachment-0001.obj
 
-- next part --
A non-text attachment was scrubbed...
Name: lamno3.relax.in
Type: application/octet-stream
Size: 4303 bytes
Desc: not available
Url : 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/830f52ea/attachment-0002.obj
 

[Pw_forum] Calculation of spin polarisation in odd electrons in PWscf

[Shaptrishi Sharma]

Hi QES user
I want to know how does thecalculation ofspin polarisation in odd number of
electrons been done in Quantum espresso, as I have odd number of eelctrons
in my system which is semiconductor.

Thanks

Shaptarishi
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/16c69a41/attachment.htm
 

[Pw_forum] runtime error

[Paolo Giannozzi]

On Saturday 08 August 2009 19:41, Aritz Leonardo wrote:

> Any idea where to look for solutions?

first of all you should look into the output, preferrably
by running interactively (no redirection to a file). If no
indication comes from the output: there is something 
wrong in your compiler/librararies. Since g95 is very 
good and very reliable, I would say that there is 
something wrong with your mpi libraries

Paolo
-- 
Paolo Giannozzi, Democritos and Udine University

[Pw_forum] References for damped dynamics structural minimization?

[Paolo Giannozzi]

On Tuesday 04 August 2009 16:14, J. J. Ramsey wrote:

> I've been trying to find references for the damped quick-min Verlet
> algorithm used by PWscf in structural relaxation (when calculation =
> 'relax' and ion_dynamics = 'damp')

if you find anything, please let us know

P.
-- 
Paolo Giannozzi, Democritos and Udine University

[Pw_forum] scf

[Eyvaz Isaev]

I am not sure convergence may depend on degauss value. To avoid this choice you 
can use  "tetrahedra' method.

Did you check your system by XCrysDen?

Bests,
Eyvaz.

---
Prof. Eyvaz Isaev, 
Theoretical Physics Department, Moscow State Institute of Steel & Alloys, 
Russia, 
Department of Physics, Chemistry, and Biology (IFM), Linkoping University, 
Sweden 
Condensed Matter Theory Group, Uppsala University, Sweden 
Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com


--- On Fri, 8/14/09, ?lvaro Alves  wrote:

> From: ?lvaro Alves 
> Subject: Re: [Pw_forum] scf
> To: "PWSCF Forum" 
> Date: Friday, August 14, 2009, 1:43 AM
> Hello.
> I am doing a scf calculation system and uses degauss = 0.003
> and
> needed change it? for 0001. With this new value, the
> calculation does
> not converge. What may be happening?? 
> 
> 
>  Regards, Alvaro
> 
> A. S. Santos
> 
> --- Em qui, 13/8/09, ?lvaro Alves
>  escreveu:
> 
> De: ?lvaro Alves 
> Assunto: [Pw_forum] scf
> Para: "pw" 
> Data: Quinta-feira, 13 de Agosto de 2009, 17:30
> 
> Hello I am doing a scf calculation system and
> uses degauss = 0.003 and
> needed change it paar 0001. With this new value, the
> calculation does
> not converge. What may be happening? 
> 
> 
>  Regards, Alvaro
> 
> 
> 
> 
> 
> 
>   Veja quais s?o os assuntos do momento
> no Yahoo! + Buscados: Top
> 10 - Celebridades
> - M?sica
> - Esportes
> -Anexo incorporado-
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 
> 
> 
> 
>   Veja quais s?o os assuntos do momento
> no Yahoo! + Buscados: Top
> 10 - Celebridades
> - M?sica
> - Esportes
> -Inline Attachment Follows-
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


  

[Pw_forum] scf

[Eyvaz Isaev]

Dear Alvaro,

--- On Fri, 8/14/09, spmmal  wrote:

>> Hello I am doing a scf calculation system and uses? degauss = 0.003 
>> and? needed change it paar 0001. With this new value, the calculation >> 
>> does? not converge. What may be happening?
 ?
> Is there any general rule of setting parameters to make the
> SCF calculation? converge. Thanks.

First of all, you shoud follow general Netiquette accepted in this forum: 
provide your affiliation.

Then the best choice (or a general rule) is the use of default parameters.
See /Doc/PW.txt file and example files. Change default value if you know what 
you are doing.

Bests,
Eyvaz.

---
Prof. Eyvaz Isaev, 
Theoretical Physics Department, Moscow State Institute of Steel & Alloys, 
Russia, 
Department of Physics, Chemistry, and Biology (IFM), Linkoping University, 
Sweden 
Condensed Matter Theory Group, Uppsala University, Sweden 
Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com




  

[Pw_forum] scf

[Álvaro Alves]

Hello. I am doing a scf calculation system and uses degauss = 0.003 and
needed change it? for 0001. With this new value, the calculation does
not converge. What may be happening?? 


 Regards, Alvaro

A. S. Santos

--- Em qui, 13/8/09, ?lvaro Alves  escreveu:

De: ?lvaro Alves 
Assunto: [Pw_forum] scf
Para: "pw" 
Data: Quinta-feira, 13 de Agosto de 2009, 17:30

Hello I am doing a scf calculation system and uses degauss = 0.003 and
needed change it paar 0001. With this new value, the calculation does
not converge. What may be happening? 


 Regards, Alvaro






  Veja quais s?o os assuntos do momento no Yahoo! + Buscados: Top 10 - 
Celebridades - M?sica - Esportes
-Anexo incorporado-

___
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



  

Veja quais s?o os assuntos do momento no Yahoo! +Buscados
http://br.maisbuscados.yahoo.com
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/5b6243c3/attachment.htm
 

[Pw_forum] compile error, pathscale compiler

[Jess Kondor]

hello all,

  I use pathscale compiler (3.2) to  compiler q-e, v.4.1. I got this
error message:

mpif90 -O2 -OPT:Olimit=0 -ftpp -march=opteron -D__EKO -D__FFTW3
-D__MPI -D__PARA  -I../include  -I./  -I../Modules  -I../iotk/src
-I../PW  -I../PH -c mp_base.f90

CALL synchronize( gid )
 ^
pathf95-700 pathf90: ERROR BCAST_REAL, File = mp_base.f90, Line = 62,
Column = 14
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

CALL synchronize( gid )
 ^
pathf95-700 pathf90: ERROR BCAST_INTEGER, File = mp_base.f90, Line =
111, Column = 14
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

CALL synchronize( gid )
 ^
pathf95-700 pathf90: ERROR BCAST_LOGICAL, File = mp_base.f90, Line =
157, Column = 14
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR REDUCE_BASE_REAL, File = mp_base.f90, Line
= 252, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR REDUCE_BASE_INTEGER, File = mp_base.f90,
Line = 383, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR REDUCE_BASE_REAL_TO, File = mp_base.f90,
Line = 484, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR REDUCE_BASE_INTEGER_TO, File = mp_base.f90,
Line = 572, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR PARALLEL_MIN_INTEGER, File = mp_base.f90,
Line = 660, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR PARALLEL_MAX_INTEGER, File = mp_base.f90,
Line = 756, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR PARALLEL_MIN_REAL, File = mp_base.f90, Line
= 851, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

  CALL synchronize( comm )
   ^
pathf95-700 pathf90: ERROR PARALLEL_MAX_REAL, File = mp_base.f90, Line
= 947, Column = 8
  The intrinsic call "SYNCHRONIZE" is being made with illegal arguments.

pathf95: PathScale(TM) Fortran Version 3.2 (f14) Thu Aug 13, 2009  10:14:01
pathf95: 1277 source lines
pathf95: 11 Error(s), 0 Warning(s), 0 Other message(s), 0 ANSI(s)
pathf95: "explain pathf95-message number" gives more information about
each message
make[1]: *** [mp_base.o] Error 2

Unfortunately, I don't know fortran very well, so I cannot figure it
out what is wrong. BTW, pgi compiler can compile this module.

-- 
=
Jess Kondor

NSTI, PhD
=

[Pw_forum] scf

[Álvaro Alves]

Hello I am doing a scf calculation system and uses degauss = 0.003 and
needed change it paar 0001. With this new value, the calculation does
not converge. What may be happening? 


 Regards, Alvaro





  

Veja quais s?o os assuntos do momento no Yahoo! +Buscados
http://br.maisbuscados.yahoo.com
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090813/883fbee6/attachment.htm
 

[Pw_forum] IBM-AIX vs ifort

[Paolo Giannozzi]

On Wednesday 05 August 2009 12:20, Bipul Rakshit wrote:

> I am doing a relax calculation of LaMnO3, the linux doing
>  Entering Dynamics:iteration =2
>
> while the AIX-IBM is doing
>  Entering Dynamics:iteration =12
>
> Can any body suggest why it is happening?

no, nobody can suggest anything, on the basis of
such a confused and obscure message

Paolo
-- 
Paolo Giannozzi, Democritos and Udine University

[Pw_forum] DOS and pDOS parallelization

[Andrea Ferretti]

Dear Madhura,

>
>  Thanks for the explanation.
>  I had always thought that DOS code also requires wavefunction files; and
> that is why one needs to give the same prefix and outdir path as in scf
> calculations. If code does not need wavefunctions, which files it uses
> for DOS calc.s?

the code needs mainly the eigenvalues, which are stored in
$prefix.save/K?/eigenval.xml  files,
where K? (i.e. K1, K2, etc ) correspond to each kpt.


>  As per suggestion, I have run PDOS calc.s with the same number of pools
> and processors (in case of B) and got results. That means PDOS code is
> parallelized with pool implementation, but not DOS.
>

yes you are right,
this means that pDOS is parallelized over G vectors and pools, while
DOS calculation is not parallelized but is anyway compatible with
the parallel environment (i.e. it does not crash if run parallel)

andrea

>  Thanks once again,
>  Madhura.
>
> >
> > Dear Madhura,
> >
> > the issue of using the same parallelism scheme (# of processors, # of
> > pools, etc) in postprocessing calculations is mainly concerned with the
> > need to read wfcs. As you wrote, if you specify wf_collect = .TRUE. you
> > completely remove this issue (you can read wfcs whatever the
> > parallelization
> > scheme you used to produce them).
> >
> >>From this point of view, the main difference between DOS and pDOS
> > calculations is that the former does not need to read wfcs, while the
> > latter does.
> >
> > Moreover, the taks the DOS program has to perform is so inexpensive that
> > it is performed by a single processor even if the code is run in parallel.
> > This is consistent with your observation at (ii).
> >
> > In the case of pDOS (i.e. when running projwfc.x) if you do not collect
> > wfcs you should run with the same # of processors and the same pool
> > parallelism. Otherwise, as you experienced, you get a davcio error
> > (some of the wfcs files were not found).
> > According to me, anyway, projwfc should work with pools if the above
> > conditions are fulfilled.
> >
> > hope it helps
> > andrea
> >
> >
> >>
> >>  There have been many discussions in this forum about parallelization of
> >> DOS and projected DOS codes. However, some of the points were not clear
> >> to
> >> me, so I have performed some scf calculations for a very simple system
> >> using 8 processors and with flag wf_collect 'false'. I have performed
> >> same
> >> calculations once without use of pools (A) and then using 2 pools (B).
> >> Then using theses wavefunctions I have performed DOS and PDOS
> >> calculations. Following is the summary of results and my interpretation:
> >>
> >> i) Pool parallelization is not implemented for both these codes.
> >>
> >> ii) DOS calculations: for both the cases A and B, one can calculate DOS
> >> with the same no. of processors (= 8) and the results match within
> >> numerical errors; even though for case B, the wavefunctions were
> >> obtained
> >> with pooling and DOS without it.
> >>  For case A, even if I use 4 processors I get identical results as when
> >> I
> >> use 8 processors. (Note, I have not checked with less no. of proc.s for
> >> B).
> >> => The condition that we need the same no. of processors and pools as
> >> were
> >> used in scf calculations is not necessary; and it is possible to get DOS
> >> results even with wavefunctions generated with pool parallelization.
> >>
> >> iii) PDOS calculations: This can be calculated only in case of A and
> >> using
> >> the same no. of processors. If I use wavefunctions generated in case B
> >> or
> >> less no. of processors (= 4) with A wavefn.s then I get "davcio" error.
> >> =>
> >> For PDOS calculations, one cannot use wavefunctions generated with pool
> >> parallelization unless wf_collect flag is set to 'true' for scf calc.s;
> >> also one has to use the same no. of proc.s as were used for scf calc.s
> >> to
> >> get projected DOS.
> >>
> >> Now my question is are these interpretations correct? Or they may change
> >> for some other system?? Do I need to do some more checks to ascertain
> >> these? If yes, what sort of tests?
> >>
> >> Thanks for reading this long mail patiently, but I need to clarify on
> >> these points before I can start with bigger systems.
> >> Sincerely,
> >> Madhura.
> >>
> >>
> >>
> >
> > --
> > Andrea Ferretti
> > MIT, Dept Material Science & Engineering
> > bldg 13-4078, 77, Massachusetts Ave, Cambridge, MA
> > Tel: +1 617-452-2455;  Skype: andrea_ferretti
> > URL: http://quasiamore.mit.edu
> >
> > Please, if possible, don't send me MS Word or PowerPoint attachments
> > Why? See:  http://www.gnu.org/philosophy/no-word-attachments.html
> >
> >
> > ___
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
> >
>
>
>

-- 
Andrea Ferretti
MIT, Dept Material Science & Engineering
bldg 13-4078, 77, Massachusetts Ave, Cambridge, MA
Tel: +1 617-452-2455;  Skype: 

[Pw_forum] negative (alpha)^2*F(w)

[Ali Tavana]

Dear QE users,

I have a question about e-ph lambda, calculated by "matdyn" program. I've found 
that for some frequencies the quantity "(alpha)^2*F(w)/w" or "dos_a2F(w)/w" can 
be negative, most probably because of the negative (imaginary) frequencies (I 
can't believe a2F(w) become negative!). In "matdyn" program these negative 
values are considered in lambda calculations. Should one remove these negative 
values as these imaginary frequency modes can never exist in the system? then 
where are those phonon branches? I think negative values of frequency can show 
in some manner more strong electron structure interaction, so what do you 
suggest one should do?

regards,
Ali




  

[Pw_forum] DOS and pDOS parallelization

[Madhura Marathe]

 Dear all,

 There have been many discussions in this forum about parallelization of
DOS and projected DOS codes. However, some of the points were not clear to
me, so I have performed some scf calculations for a very simple system
using 8 processors and with flag wf_collect 'false'. I have performed same
calculations once without use of pools (A) and then using 2 pools (B).
Then using theses wavefunctions I have performed DOS and PDOS
calculations. Following is the summary of results and my interpretation:

i) Pool parallelization is not implemented for both these codes.

ii) DOS calculations: for both the cases A and B, one can calculate DOS
with the same no. of processors (= 8) and the results match within
numerical errors; even though for case B, the wavefunctions were obtained
with pooling and DOS without it.
 For case A, even if I use 4 processors I get identical results as when I
use 8 processors. (Note, I have not checked with less no. of proc.s for B).
=> The condition that we need the same no. of processors and pools as were
used in scf calculations is not necessary; and it is possible to get DOS
results even with wavefunctions generated with pool parallelization.

iii) PDOS calculations: This can be calculated only in case of A and using
the same no. of processors. If I use wavefunctions generated in case B or
less no. of processors (= 4) with A wavefn.s then I get "davcio" error. =>
For PDOS calculations, one cannot use wavefunctions generated with pool
parallelization unless wf_collect flag is set to 'true' for scf calc.s; 
also one has to use the same no. of proc.s as were used for scf calc.s to
get projected DOS.

Now my question is are these interpretations correct? Or they may change
for some other system?? Do I need to do some more checks to ascertain
these? If yes, what sort of tests?

Thanks for reading this long mail patiently, but I need to clarify on
these points before I can start with bigger systems.
Sincerely,
Madhura.


-- 
Madhura Marathe,
PhD student, TSU,
JNCASR, Bangalore.
India.
Phone No: +91-80-22082835