two quad-core processors.
I would like to compile some processors, too ;)
But to use these 8 cores you have too compile in parallel and use mpiexec
to run siesta.
Cheers Marcel
Thank you for your helpfull comments
best regards to all.
Walter
Walter Alberto Cañón Mancisidor
Químico
Doctorado
one give k-points explicitly one by one?
As far as I know, no.
Cheers Marcel
did you try
diag.paralleloverk T ?
Regards
Marcel
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL: +49-30-314 24442 Hardenbergstr. 36
Hi
I think you have to specify which wavefunction you want
WaveFuncKPointsScalepi/a
%block WaveFuncKPoints
0.0 0.0 0.0 from to
%endblock WaveFuncKPoints
Cheers marcel
Marcel Mohr
ocal part of the psp?
Cheers Marcel
____
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL: +49-30-314 24442 Hardenbergstr. 36
FAX: +49-
have a look at the bottom of
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/list.html
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL
On Mon, 22 Sep 2008, Adrain Zhou wrote:
Dear siesta user,
I have a user id and a password provided by siesta for
downloading siesta package. But when i try to download
siesta, then it shows 'not found'
page.
Same problem here. Wanted to upgrade to 2.0.1
Marcel
dynamics)
Cheers
Marcel
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL: +49-30-314 24442 Hardenbergstr. 36
FAX: +49-30-314 27705
This was discussed at least 3 times in the last year.
Cheers Marcel
Hint: There is no switch in siesta for calculating the bulk modulus
On Tue, 10 Jun 2008, madani samah wrote:
Dear siesta users;
Please can any body tell me how to compute the bulk modulus and what are the
instructions to
f90 compiled, compile SIESTA.
Cheers Marcel
Could you give me again comments?
With kind regards,
Chol-Jun
Eduardo Anglada wrote:
Hi,
You compiled blacs and scalapack with gfortran or g77, while siesta was
compiled
with ifort (intel fortran). You shouldn't mix compilers.
Regards,
Eduar
linux) and do semiautomatically what is written in step 2.)
Cheers Marcel
Could you please explain me in details.
with regards
Sonia Mehra
- Original Message
From: Marcel Mohr <[EMAIL PROTECTED]>
To: SIESTA-L@listserv.uam.es
Sent: Wednesday, 30 April, 2008 5:26:56 PM
Subje
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/
On Tue, 6 May 2008, murali krishna wrote:
To SIESTA-L organizers,
I want to un-register from the e-mail list of SIESTA users.
Please kindly do the needful or write to me the procedure of how to un-register
Thanks all
Murali krish
olume =" outputfiles
grep "otal =" outputfiles
and you get the cell volume and total energy.
Slightly more elegant with bash programmint:
for i in output* ; do vol=`grep "olume =" $i` ; en=`grep "otal =" $i` ; echo $i
$vol $en ; done >vol_en.dat
Cheers
Marcel
Hello Roberto
I think it is just an error in your input file, e.g., you have
specified "Number of Atoms 12" and you have only entered 10 coordinates.
Maybe post it here
Cheers Marcel
On Mon, 28 Apr 2008, Roberto Sapiens wrote:
I'm gettin the following error when running
Dear Andrei, dear Vasili
thanks a lot for your shared knowledge.
Cheers Marcel
Hi!
2008/4/15, Marcel Mohr <[EMAIL PROTECTED]>:
The smaller energyshift, should in principle be better, and results in
larger basis size.
That's not true. It only has to be better for an isolate
Hi,
thank you for your reply
On Tue, 15 Apr 2008, Vasilii Artyukhov wrote:
2008/4/15, Marcel Mohr <[EMAIL PROTECTED]>:
The smaller energyshift, should in principle be better, and results in
larger basis size.
That's not true. It only has to be better for an isolated atom, who
in principle be better, and results in
larger basis size.
Which energyshift would you recomment in this case?
And can anyone explain to me, why the larger basis gives an overestimation
of the lattice constant?
Cheers
Marcel Mohr
lattice constant.
How serious should I take this overestimated lattice constant?
Kind regards
Marcel
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL: +49-30
acknowledged.
Cheers
Marcel
Marcel Mohr Institut für Festkörperphysik, TU Berlin
marcel(at)physik.tu-berlin.de Sekr. EW 5-4
TEL: +49-30-314 24442 Hardenbergstr. 36
FAX: +49-30-314 27705
688.550508
Along z the polarization is, as expected, 3 times larger. But what
happened to the other 2 directions?
It is not really important, but I would like to understand it.
Cheers Marcel
Marcel Mohr
point every time.
This makes it very difficult to achieve an unambiguous convergence to just one
(global) minimum.
Is there any way for the CG to reuse its most recent sets of parameters.
Have you tried these parameters?
MD.UseSaveCG.true.
MD.UseSaveXV.true.
Cheers Marcel
Hello Bipul
On Sat, 6 Oct 2007, bipul rakshit wrote:
hello Marcel,
I have only one doubt in the above thing. If consider i have 54 atoms in a
system. So in FC.fdf file its written
MD.FCfirst 27
MD.FClast 28
So the file.FC file contains 649 lines..
Yes
consider my calculation
Dear list-members,
I checked the publications list to find papers that use SIESTA for the
calculation of vibrational properties of nanowires, but could not find any. (Only
nanotubes).
Does anybody know work on this topic?
Cheers
Marcel Mohr
In addition your k-point grid looks strange
siesta: k-grid:0 0 8 0.500
siesta: k-grid:0 2 0 0.500
siesta: k-grid:2 0 0 0.500
maybe try something like
siesta: k-grid:1 0 0 0.000
siesta: k-grid:0 1 0 0.000
siesta: k-grid:0 0
Dear Vikas
You could delete the *DM file and restart.
However, your unit cell seems not ok?
I would say 6,0 has 24 atoms in the unit cell.
Cheers Marcel
Marcel Mohr Institut für Festkörperphysik, TU
correct me, if this is wrong. And maybe an additional question,
would that then be enough? (C with s,p and d basis functions)
Cheers Marcel
2007/7/4, Marcel Mohr <[EMAIL PROTECTED]>:
Dear Vikas, dear Andrei
to circumvent the error message and still use a pseudo with only L=2 one
cou
with L=3 too, but on other systems I do not
always follow the recommondation of SIESTA).
Regards Marcel
On Wed, 4 Jul 2007, Andrei Postnikov wrote:
On Wed, 4 Jul 2007, vikas thakur wrote:
| Dear users,
| Please try to find some time for my problem.
| it reads the input file but do not
symbol) and a default color are associated at each atom species.
Riccardo
I remember this maybe did not work in version 5.1, but for version 4.3 of
molekel it does work. I still prefer 4.3 over 5.1 ;-)
Cheers
Marcel
Dear Saswata,
and it looks like you use a huge unit cell. Is that really what you want?
(Remember, the lattice vectors are multiplied by the lattice constant)
Cheers Marcel
On Mon, 19 Mar 2007, Eduardo Anglada wrote:
Dear Saswata,
I think you don't have enough free memory.
Re
l. (I wouldnt even find 120 atoms really
big, e.g. compare with doping concentrations).
Kind regards
Marcel
Marcel Mohr Institut für Festkörperphysik, TU Berlin
[EMAIL PROTECTED] Sekr. PN 5-4
TEL: +
moved to
then.
So at last, your FC file should have 6 x atoms in cell> + 1 lines.
I hope this was not too confusing.
Regards Marcel
On Fri, 2 Mar 2007, bipul rakshit wrote:
hello siesta user,
is it possible to restart the phonon calculations???
-
Heres a
Dear Yurko
maybe have a try with "COMP_LIBS=linalg.a".
Below is my arch.make which worked fine.
Regards
Marcel Mohr
#
FC=mpif90
#
FFLAGS= -w -O3 -tpp7
FPPFLAGS= -DMPI -DFC_HAVE_FLUSH -DFC_HAVE_ABORT
FFLAGS_DEBUG= -g
LDFLAGS=-Vaxlib -static
COMP_LIBS=
RANLIB=echo
#
N
ack
BLAS=-lmkl_ia32
These are the intel mkls, aren't they? Better try your own compiled.
Regards
Marcel
Hello Rodrigo
it would help if you post your arch.make.
I guess you did not compiled blas, lapack yourself.
(it is easy, you may have a look at mail from sebastian leRoux
in the archive with a pdf howto).
Regards
Marcel
Hello
I guess Fe has a lot of k-points.
Try
diag.paralleloverK true
As it says in the manual, parallelization can be strongly dependant on the
hardware used.
Regarsd
Marcel
Marcel Mohr Institut für
needed
for the actual problem).
However, what do you mean with Ring shaped?
Good luck
Marcel
On Thu, 23 Nov 2006, Michael Shin wrote:
Hello SIESTA users
I am doing some calculations on BN(Boron Nitride) nanotube(5,5) which is
semiconductor. I am using 20 atoms with in GGA.
During the CG
lattice
parameter.
The obtained function Total energy vs. lattice parameter will have a
minimum at the relaxed position.
Good Luck
Marcel
On Thu, 23 Nov 2006, Oscar Olvera Neria wrote:
I want to calculate the lattice parameter for Pd, I
obtain the next result after I optimize the cell:
outcell
pancy?
Greetings
Marcel Mohr
Dear Sui Yang
have a look in a mail archive at sebastian le Roux, he has written a guide
for compiling all these libs.
you also need intel C- compiler
Regards
Marcel
Marcel Mohr Institut für
Dear Tao
I'm not sure it helps, but maybe you have to set the -DMPI Flag when
compiling.
Maybe you should add the flag '$(DEFS)' in the lines with big F's:
'.F.o' and '.F90.o'
Hope that helps
Regards
Marcel
PS:
I used the intel mkls and didn't
Sophisticated guess from the error message:
check your netcdf library.
Or leave it empty.
Regards
On Fri, 22 Sep 2006, Tao K wrote:
Dear Siesta user,
These days I try to compile parallel Siesta-2.0 on a AMD-Opteron
cluster, but I always fail to do this and encounter a problem:
fortcom: Erro
Dear Sen
maybe the order of LIBS is wrong in your arch.make.
It should be $SCALAPACK $BLACS $LAPACK $BLAS .
Have you compiled BLACS and SCALAPACK with mpif77 ?
Regards Marcel
Marcel Mohr Institut für
completes without
errors.
I put the correct arch.make in the attachment.
Regards
Marcel Mohr
On Thu, 7 Sep 2006, Marcos Verissimo Alves wrote:
Hi Marcel,
Maybe it's best to use lf95 for everything, even with eventual
preprocessing (that means, if some files have to be preprocessed
Hi Chun,
you can try to run some simple mpi programm, like the cpi in examples
directory in parallel, to test mpi. You also can do a ' make testing' with
a smpd running on a different node.
Regards
Marcel
On Sun, 10 Sep 2006, Marcos Verissimo Alves wrote:
Hi Chun,
Do you mea
Hi Marcel,
You should use the same compiler. If you change from one to another the
compilation
is going to be a nightmare. It is possible but really, really tricky.
Regards,
Eduardo
OK, thanks. Now I think I did use the same compiler , only I used a
precompiled ATLAS library. (The
Hello all
I am trying to compile Siesta and required packages (BLACS, scalapack)
from scratch.
However do I have to use the SAME Fortran compiler for all packages?
Or can I use GNU f77 for BLACS & scalapack and LaheyFujitsu lf95 for mpi
and SIESTA ?
Kind regards
Marcel Mohr
Hello Bipul
if you calculate some bulk material, (i dont know what sms means), the
unit cell seems to be to small, so you get an pressure outwards. Try to
increase your unit cell or use DM.VariableCell T
Marcel Mohr
deformation
potential, which is in 1st order perturbation theory proportional to
electron phonon matrix element.
Have a look in
Machon et al. PHYSICAL REVIEW B 71, 035416 2005
Regards Marcel
On Tue, 15 Aug 2006, Mousumi Upadhyay Kahaly wrote:
Dear All,
If I have full phonon dispersion as well
.
Kind Regards
Marcel Mohr
Please find below some more details on my calculations:
PAO "convergence":
==
PAO.energyshift Total Energy
meV eV
0.5 -2947.565291
1 -2947.552009
5 -2947.521170
10 -2947.445567
20 -2947.305253
40 -2947.072458
60 -2946.
should
comment on that, as I don't have any experience with it.
Greetings
Marcel Mohr Institut für Festkörperphysik, TU Berlin
[EMAIL PROTECTED] Sekr. PN 5-4
TEL: +49-30-314
Dear all,
I am looking for pseudos for Cd and Se. Preferably LDA and
non-relativistic.
None are listed in the database.
Thanks in advance
Marcel Mohr
Marcel Mohr Institut für Festkörperphysik, TU
51 matches
Mail list logo
