/~luxu
==
Eduardo Anglada writes:
Hi,
Could you please post your results and the input fdf?
The energy should converge.
Best
Eduardo On 01/11/2008, at 15:23, Xu Lu wrote:
Dear all :
I have a problem in calculating the energy of single atom. The
energy
Could you please post your V pseudo?
The mesh cutoff doesn't vary the scf, it only controls the rippling of
energy/forces.
This rippling is due to the FFT aliasing of those magnitudes
(orbs^2,neutral atom potential, core
charge for non linear core corrections) which are projected into the
Hi,
At the beginning of each SIESTA run, during the generation
of the atomic orbitals, the number of basis functions for each
species is written.
Best
Eduardo
On 03/11/2008, at 21:05, Roberto Veiga wrote:
Where in the output can I find the number of basis functions?
Roberto
From: Oleksandr
do I obtain the total number of basis
functions for a condensed system? Should I take into account more
than the atoms explicitly defined in my input?
Roberto
From: Eduardo Anglada [EMAIL PROTECTED]
To: SIESTA-L@listserv.uam.es
Sent: Tuesday, November 4, 2008 9:45:34 AM
Subject: Re: [SIESTA-L
Hi,
Could you please post your results and the input fdf?
The energy should converge.
Best
Eduardo
On 01/11/2008, at 15:23, Xu Lu wrote:
Dear all :
I have a problem in calculating the energy of single atom. The
energy will drop linearly with respect to decreasing of energy
shift. It
***
I'm fairly confused now which compiler to use for using Siesta and
Utils.
Note: Siesta works fine now :)
Thanks,
Regards,
From: Eduardo Anglada [EMAIL PROTECTED]
To: SIESTA-L@listserv.uam.es
Sent: Thursday, October 30, 2008 9:55:07 AM
Subject: Re: [SIESTA-L
Hi,
As far as I know It is not being distributed.
Best,
Eduardo
On 29/10/2008, at 12:48, José Eduardo Padilha de Sousa wrote:
Hi All.
In a recent article about SIESTA (J. Phys.: Condens. Matter 20
(2008) 064208), I
read that a multigrid solver for the Poisson equation on the grid
has been
Hi,
Try with:
a.out input
Best,
Eduardo
On 30/10/2008, at 5:18, Руслан Жачук wrote:
Dear SIESTA users,
I am using OpenSUSE Linux 10.3 and installed following packages to
compile SIESTA utilities:
1) gcc-fortran (4.2) //The system GNU Compiler
2) gcc42-fortran (4.2.1_20070724) //The GNU
Hello
You have to use the same compiler for the utils and for siesta.
Could you please post the error message?
Best,
Eduardo
On 29/10/2008, at 18:17, Johnny Dry wrote:
Hello Siesta users,
I'm using:
* Mandriva Linux 2008
* g95 compiler
* Siesta 2.0.1
I've compiled these packages:
* Siesta
Hola Roberto,
Then the problem is that your compiler does buffering I/O.
As the buffer isn't full there is no flush of the communication
channel and the slave siesta waits forever.
Almost all the compilers have an option which turns off
the buffering.
Best,
Eduardo
On 22/10/2008, at 12:50,
Dear Alexandre,
Why do you fix the radius of all the channels to the same value?
In principle the matching radius is different for each angular
momentum channel.
Best regards,
Eduardo
On 15/10/2008, at 16:00, Alexandre Lebon wrote:
Dear Siesta users,
I am meeting some difficulty at
On 11/10/2008, at 9:52, Subhra Kulshrestha wrote:
Dear users,
I am also facing the same problem in computing the band structure of
f-band materials. They are semiconducting but the band structure
calculated from LDA, LSDA, GGA, GGA+spin gives metallic behaviour.
I want to get it
Dear Karim,
It is possible, but you have to construct the program yourself.
Best,
Eduardo
On 11/10/2008, at 14:44, karim rezouali wrote:
Dear SIESTA users,
Is it possible to calculate the following parameters using the
SIESTA code?
1. Pair correlation function
2. the self diffusion
Hi,
You have to compile plrho exactly (with the same compiler and options)
the same as the siesta you are going to use.
If you mix compilers and/or options it won't work.
Best,
Eduardo
On 07/10/2008, at 14:59, zhiyong wang wrote:
hi,all siesta users:
I have compiled the plrho in my
Hi,
David is correct, you have to decrease the electronic temperature.
If you encounter any difficulties send me a message.
Best regards,
Eduardo
On 03/10/2008, at 19:40, David Strubbe wrote:
I think the LDOS is broadened by the electronic temperature, so maybe
you have to decrease that.
Hi Arun,
This problem happens from time to time. It's quite frustrating because
I haven't been able to reproduce it, so I don't have any clue.
What happens if you restart?
Does the output include the cell vectors of the last step?
Maybe they can be found in the XV file.
The crash happens during
Hi Kamaram!
Do you need so many planes? The calculation is going to be very long!
If you need the LDOS you should include the following option in the
input fdf:
%block LocalDensityOfStates
Emin Emax units
%endblock LocalDensityOfStates
The Emin and Emax (Emin Emax) specify the
Dear Adrian,
We are really sorry, we're trying to solve this issue.
If you can't wait I will send it to you (or anybody else).
Best regards,
Eduardo
On 30/09/2008, at 5:24, Adrain Zhou wrote:
Dear all,
I can not download SIESTA code from
Hi,
Are you able to run (successfully) scalapack and blacs tests?
Maybe the problem has nothing to do with siesta, it looks like
a communication problem.
Regards,
Eduardo
On 23/09/2008, at 4:30, M Bharat Kumar wrote:
Hello All,
When trying to run simulations using siesta on our new cluster,
Dear Nidhi,
On 24/09/2008, at 8:50, Nidhi Sharma wrote:
Dear Users,
I have computed the high pressure properties and electronic
structure calculations of a semiconducting heavy earth compounds
with LDA and non-relativistic, alongwith polarization orbitals :
perturbative polarization.
Dear Chol-Jun Yu,
Siesta can do this kind of molecular dynamics simulations.
In the input fdf file you should include the following options:
WriteMDhistory
WriteMDXmol
At the end of the simulation there will be two output files:
SystemLabel.MD - contains positions and velocities at every time
On 02/10/2008, at 17:58, Bharat wrote:
Hi Eduardo,
Thanks. scalapack and blacs work fine.
Anyhow I figured its a problem with mvapich2. With openmpi, I could
run the problematic
input files successfully.
Great!!
On Thu, 02 Oct 2008 01:52:42 -0600, Eduardo Anglada [EMAIL PROTECTED
Hi,
I'm really sorry but the PBE0 functional is not implemented in SIESTA.
Regards,
Eduardo
On 01/08/2008, at 7:54, Adrain Zhou wrote:
Dear all,
Is there anybody has experience with hybrid functional
(PBE0) calcualtion? Could you please tell me how is
the performance? Do I use only PBE atom
Hi Tom,
There is a problem in Makefile or arch.make files,
the make program doesn't know how to compile a given
file, and it call the c compiler without any input.
Regards,
Eduardo
On 10/08/2008, at 4:08, Thomas Sadowski wrote:
Hello all,
I am attempting to compile parallel siesta on an
Hi Lakee,
I think you have a problem with you openmpi setup, it tries to use
infiniband,
but it can't find any network cards so it uses standard ethernet.
What I recommend is:
1) Check the tests and examples of openmpi.
2) Compile and test Blacs. If the test fail SIESTA will fail, so don't
Hi,
I've asked a person who runs siesta in Blue Gene and this is her
arch.make.
She had to compile scalapack by hand.
SIESTA_ARCH=XLF 32bits PARALLEL
#
FC= mpixlf90 #xlf90_r
#
FFLAGS_DEBUG= -g
FFLAGS= -O5 -qarch=440d -qtune=440
COMP_LIBS=
RANLIB=echo
#
NETCDF_LIBS=
NETCDF_INTERFACE=
Hi,
Please take a look at this previous posting by Javier Junquera. He has
written
a nice review about the subject
http://www.mail-archive.com/siesta-l@listserv.uam.es/msg00627.html
Best regards,
Eduardo
On 18/07/2008, at 20:34, zubaer wrote:
Hi,
I wanted to calculate the workfunction
of the zigzag profile?
Luis
From: Siesta, Self-Consistent DFT LCAO program, http://www.uam.es/siesta
[mailto:[EMAIL PROTECTED] On Behalf Of Eduardo Anglada
Sent: Tuesday, July 15, 2008 7:35 AM
To: SIESTA-L@listserv.uam.es
Subject: Re: [SIESTA-L] Electric field
Hello Luis,
Siesta solves
Hello Luis,
Siesta solves the Poisson equation using ffts, so the zig-zag
potential due the electric field is allowed in
the directions where there is vacuum. But siesta doesn't introduce a
fictitious dipole charge (as far as I know).
Maybe if you change the cell geometry and the values of
Hi,
You can find the translation of abinit's pseudos into siesta psf
format here:
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/periodictable-intro.html
Regards,
Eduardo
On 13/07/2008, at 14:56, Tafoughalt Mohand Akli wrote:
Dear SIESTA users,
I need a pseudopotential
.
Thanks
-Kam
Eduardo Anglada [EMAIL PROTECTED] wrote:
Hi,
This error also happens with intel fortran compiler.
Kamaram, could you send me a copy of the fdf and pseudos?
Also I need the compilation options you used.
Regards,
Eduardo
On 27/06/2008, at 19:18, Marcos Verissimo Alves wrote:
Hm
written / Lei ha scritto / Você escreveu...
Eduardo Anglada
Hi,
That's the problem, there is no ultimate systematic scheme.
My recommendation:
Start optimizing the energyshift parameter. Once the
energy is converged with respect to this parameter then you can
optimize the splitnorm of the
multiple
Hi,
This error also happens with intel fortran compiler.
Kamaram, could you send me a copy of the fdf and pseudos?
Also I need the compilation options you used.
Regards,
Eduardo
On 27/06/2008, at 19:18, Marcos Verissimo Alves wrote:
Hm, this seems to be a complicated error. Googling malloc():
Dear Ebrahim,
I'm really sorry but it is not distributed.
Regards,
Eduardo
On 06/06/2008, at 13:09, eb na wrote:
Dear Siesta community,
could anybody tell me if the SIC correction which has been
implemented and reported in PRB 75, 045101, is already implemented
in any Siesta distributed
Hi,
Try to reduce the compilers optimization level, it seems to be very
aggressive.
Regards,
Eduardo
On 05/06/2008, at 11:14, Sharat Chandra wrote:
Hi
I am using version 2.01 with all the updates and I am trying to run
CG minimization on a large system with 720 atoms (Fe, Ti and C) a
Dear Reza,
There is a bug in siesta, it can't calculate a system with only 1
orbital ...
Regards,
Eduardo
On 04/06/2008, at 2:28, reza behnam wrote:
Dear Edan,
Thank you for your advises. According to Eduardo`s and your advises
I set the PAO.Basis as DZP and it worked. But why?just
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,
Eduardo
On 28/05/2008, at 9:11, Chol-Jun Yu wrote:
Hello
Dear Andrei,
Siesta uses the species number in order to identify
the different species, there is no subrutine which lets you identify
a species by it's label so it's quite safe to use the same label
for several species. But it can be very error prone, so I recommend
to have different labels for
Dear Swaroop,
Which are the values of the mesh in the previous CG steps?
Are they very different? Maybe we can guess what is going on.
Regards,
Eduardo
On 23/05/2008, at 14:24, M.Sairam Swaroop wrote:
Dear Eduardo
We have compiled siesta with a 64 bit compiler and we have allowed
for a
Clearly there is something wrong going on.
Can you send me (privately) the hole output
(including the output files with coordinates and
cell)
Best,
Eduardo
On 23/05/2008, at 18:26, M.Sairam Swaroop wrote:
Dear Eduardo
After you mentioned i noticed tha the mesh does not change ... i have
Dear Roberto,
Maybe what you are seeing is the rippling of the energy/forces due to
the aliasing (wrap around errors) of the fft.
The orbitals (squared) and neutral atom potential are projected into
the grid so if you change them this error can introduce
noise in the convergence of the total
On 21/05/2008, at 12:53, Subhra Kulshrestha wrote:
Dear Users,
I have computed the optical properties of Si by using two program
input.f and optical.f in the directory siesta-2.0.1/Util/Optical and
the program is successfully run for Si but as I run the calculation
for LaAs the file
Dear Philipp,
Currently we are working with pathscale for a solution
of your problem. Once we know how to compile siesta
with the latest version of the pathscale compilers I will
provide the arch.make. For the other versions of pathscale
compiler use the arch.make provided by Pablo Aguado.
Best
Hi,
For the mesh cutoff change the meshcutoff variable from the default of
100 Ry to
500 in steps of 100 Ry, if the energy converges (also take a look at
the forces) you
are done with the grid. If it is not converged you should continue
increasing the
mesh cutoff. If you reach a very huge
I'm sorry, I should have answered before!
I think that your pseudocode is right. That is
what siesta does in order to obtain the position of the
Fermi level.
Best regards,
Eduardo
On 02/05/2008, at 20:03, David Strubbe wrote:
Ebrahim,
No I never received any response, but I recently
On 03/05/2008, at 12:39, Nidhi Sharma wrote:
Hi to all,
I am trying to compute the energy for different lattice constant to
get the E vs V graph (in B1 phase of Smte using LDA). For this I
have selected the range from 5.5 to 7.5 Ang in steps of 0.05. After
Murnaghan fit optimize
Hi,
If you are using the latest version of siesta you should use another
diagonalization scheme. Try changing the following options (they are
fdf booleans: .true. or .false. )in your fdf:
Diag.AllInOne(default false, change to true)
DivideConquer (default true, change to false)
Hi,
Yes it is possible, but you should write your own constraint subroutine.
Take a look at the example in Src/constr.f
Regards
Eduardo
On 06/05/2008, at 10:52, eb na wrote:
Hello dear Siesta community,
How can I optimize the atomic coordinates and cell sizes while
keeping the cell
Hi Ali,
In the webpage there are LDA GGA pseudos, but I have no basis sets.
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/periodictable-intro.html
Regards,
Eduardo
Hi Roberto,
I attach the output of Bromine pseudo generation using the Fritz
Habber Institute (fhi) code.
The input was taken from Abinit's LDA/GGA databases.
The translation of the output of fhi to siesta leads to ghost states
(siesta uses a different definition
of the local part of the
Hi,
Why are you using the split gauss option? Have you checked the
resulting gaussians?
Do they decay to zero?
Regarding the convergence, if the Harris Energy is converged, you
could try to stop the
calculation and restart it (reading the coordinates and density
matrix). It many
Hi!
This message is not related to the pseudo of La! So please use a new
subject line
Can we see the arch.make? It contains all the compiler, libraries
related stuff
Regards,
Eduardo
On 10/04/2008, at 14:13, Pablo A. Denis wrote:
Dear Siesta user,
I
is possible although it contains 10
electrons.
This is a misprint in the head line of the Te pseudo. It was
generated
with 4d10 in the core and 5d as valence states. (Ask Eduardo
Anglada).
When I use this file then we get results but band gap in B1 phase
is ~10eV which is quite far from
.
Best,
Eduardo
On 12/03/2008, at 18:22, Eduardo Anglada wrote:
Dear Users of Siesta,
There is a new collection of pseudopotentials available for SIESTA!
I have translated the LDA/GGA pseudos in the Fritz-Haber-Instute
(FHI) format
of ABINIT's database found in http://www.abinit.org/Psps
Dear Users of Siesta,
There is a new interface to the archives of this mailing list:
http://www.mail-archive.com/siesta-l@listserv.uam.es/
It's much more easier to use than the current version and
it is english.
The siesta team would like to thank the www.mail-archive.com
for storing all the
On 02/04/2008, at 15:37, lan haiping wrote:
should we subscribe to it again ?
No, it is a copy of all the previous messages.
Now it is easier to look for useful info, the search
feature is much more useful.
Eduardo
Best
On Wed, Apr 2, 2008 at 12:47 PM, Eduardo Anglada [EMAIL PROTECTED
Hi,
You should specify:
MD.TypeOfRun CG
(conjugate gradients relaxation)
MD.NumCGSteps 100
(a maximum of 100 steps will be calculated.
Of course if the forces are relaxed before the
100 steps are calculated then siesta will stop)
The relavant section of the manual is:
Dear Sonia,
It is discribed in:
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/manual-2.0/node23.html
Regards,
Eduardo
On 01/04/2008, at 13:21, Sonia Mehra wrote:
der siesta users,
Can anyone tell me that how to control the bandlines i.e.
what type of changes can be
Dear Sonia,
There are several talks in:
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/tutorials.html
(you can find the talks and the exercises)
which describe siesta basis sets.
Also maybe you can post which is the element
you are interested in.
Regards,
Eduardo
On 28/03/2008,
it, but the number of 3d
electrons in the Psf file was zero (3d 00.00). Is this a problem?
Can I simply add 7 electrons to 3d orbital of the downloaded psf
file?
Thanks,
Reza
- Original Message
From: Eduardo Anglada [EMAIL PROTECTED]
To: SIESTA-L@listserv.uam.es
Sent: Tuesday, March 25, 2008 5:15
Dear Reza,
Which one did you download?
Best,
Eduardo
Hi,
Maybe the line with the number of spices is missing?
post your input so we can take a look.
Regards
Eduardo
On 15/03/2008, at 15:00, Kamaram Munira wrote:
I have successfully run Siesta in serial mode till now. Recently, I
am trying to run the parallel version and I get a error message
Dear Andrei,
On 13/03/2008, at 20:36, Andrei Postnikov wrote:
Many thanks for Eduardo Anglada for uploading the ABINIT pseudos!
Just two related questions:
1. Does somebody know whether there s a way to see from the .psf
file if the core correction is included, and with which radius
Dear Users of Siesta,
There is a new collection of pseudopotentials available for SIESTA!
I have translated the LDA/GGA pseudos in the Fritz-Haber-Instute (FHI)
format
of ABINIT's database found in http://www.abinit.org/Psps/?text=psps so
now
they can be used with SIESTA.
The database can
Hi,
Your system lacks the basic development utils needed to build a library.
Search the documentation and install them.
Regards,
Eduardo
On 01/02/2008, at 12:58, vikas sharma wrote:
I get the following error when run the make command in Src directory.
(cd fdf ; make module)
ar cru libfdf.a
Hi Roberto,
I think your only option is to use the DM.InitSpinAF and DM.InitSpin
block.
Take a look at the manual in:
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/
manual-2.0/node18.html#2854
Regards,
Eduardo
On 09/10/2007, at 22:03, Roberto Veiga wrote:
Hi:
I'm gonna do
Dear Jingbin,
The wsxm software reads directly any output of siesta. It's designed
for the RHO and LDOS, but the format is the same
for all the output of siesta, maybe you have to rename the other files.
wsxm is free and can be found in:
http://www.nanotec.es/
It only works in MS-windows.
Hi,
Siesta is different and I think more powerful: you obtain the bands
in a single run.
The calculation is performed with the desired accuracy (you specify
the k-sampling
with the MonkhorstPack block or the kgrid_cutoff parameter) and at
the end
of the calculation if there is a BandLines
Begin forwarded message:
From: Eduardo Anglada [EMAIL PROTECTED]
Date: 8 de octubre de 2007 11:16:41 GMT+02:00
To: Edgar Martinez Guerra [EMAIL PROTECTED]
Subject: Re: [SIESTA-L] How to define a ion?
Dear Egar,
You should create a new pseudo. In the basis block you control the
size
Hi!
remove the -L infront of the library path:
ifort plrho.f -L/usr/X11R6/lib/libX11.a /share/apps/pgplot/
libpgplot.a -o plrho
If the pglopt library was not compiled with ifort then you should
include the
other compiler libraries.
Regards,
Eduardo
On 30/06/2007, at 8:51, Cherry Y. Yates
Dear Zoya,
Siesta is designed for very large systems where, usually, the number
of symmetries is very small, so it can't use them
Regards,
Eduardo
On 13/06/2007, at 9:05, Zoya Shah wrote:
Dear everyone
I just have a question regarding symmetries. Is it possible to use
the advantage of
Dear Helean,
Try to increase the meshcutoff to something more reasonable: 100-150 Ry
Regards,
Eduardo
On 02/06/2007, at 10:47, Mu J. Helien wrote:
Dear Siesta users,
I compiled the parallel version of siesta code sucessfully, but it
seems that it can't run normally with GGA method.
For the
Dear C. H. Hu,
The Prefactor could be around 150 and the Inner radious could be 3/4
of the cutoff radious.
Regards,
Eduardo
On 14/05/2007, at 9:03, Chaohao Hu wrote:
Dear Siesta users,
How to set up the PrefactorSoft and InnerRadsoft inputing
parameters when considering the new soft
This is José M. Soler answer to this question:
Dear Nichols:
The basis set and the pseudopotentials are completely
independent things:
- If you have no l=d pseudo projector, but you use l=d basis
functions, they will feel only the local part of the
pseudopotential, exactly like in a PW program
Hi Nichols,
The grid used for the calculation of the two center integrals
(calculated in matel.f) is fixed to 2000Ry,
while the one used for the three center integrals is set by the user
using MeshCutoff. They are
completely independent.
My question is about SIESTA and how the MeshCutoff
Dear Murali,
Dear all,
I have been trying to compile Siesta using f95 compiler and was not
successful. Recently, i received help from one of the users to use
g95 or gfortran compiler. As i used it, i could compile it
successfully but when i started to run some programs that are
Dear Neil,
In order to optimize geometry you must use the variable
MD.TypeOfRun CG
The Zmatrix is a way (there are others) to provide the atomic
positions.
The relevant part of the siesta manual is:
http://www.uam.es/departamentos/ciencias/fismateriac/siesta/
manual-2.0/node20.html
Regards,
Dear Saswata,
I think you don't have enough free memory.
Regards,
Eduardo
Dear You Lin,
Yes, they are bohr/fs. The XV uses the internal units of siesta: Ry,
Bohr, fs.
Regards,
Eduardo
On 08/03/2007, at 16:29, You Lin wrote:
Hi, dear siesta developers:
I'm trying to make change to .XV file. It looks like there's no
document about its format. Then I made a
Dear You Lin,
Siesta is really efficient treating empty space, so you should
increase the vacuum size
until there is no change in, for example, the total energy.
Regards,
Eduardo
On 03/03/2007, at 19:44, You Lin wrote:
Dear siesta collegues:
My simulated system is molecular system with a
On 06/03/2007, at 10:30, bipul rakshit wrote:
hello siesta users,
can any body tell me what is the exact way of using the gen basis
programme.
here i am sending the Sn.psf and Sn.fdf file
I use the script as follows:
i am in the directory of Sn
sh ../gen-basis.sh Sn.fdf Sn.psf
and then i got
Dear Nilolaos,
The numerical atomic orbitals can be found in the ORB* files which
siesta produces
in the directory where it is run. The radial part of the orbitals (phi
(r)) are stored as r, phi(r)/r**l where l is the quantum
number of each one. The ORB* files are named following this
Dear Lindsay,
In which platform and which compiler are you using?
Regards,
Eduardo
On 07/03/2007, at 16:36, Lindsay Shuller wrote:
Dear Siesta users,
I am new to the Siesta program and have compiled the version
siesta2.0. I'm trying to do some test runs (following the manual),
but I
On 13/02/2007, at 3:22, Andrey V. Semichaevsky wrote:
Hi Andrey,
You're right, phiatm and rphiatm are the interface to the basis
functions.
Am I correct to say that phiatm.f is going to give me an
implicitly angularly-dependent basis function if I call this
function for
various values
On 09/02/2007, at 13:29, Rainer WILCKE wrote:
Hello,
Those tests aren't ment to be run in parallel, they are for
the serial version.
It's true that siesta should give an error message.
Regards,
Eduardo
Hello,
I am trying to create a MPI-version of SIESTA 2.0.1 at the ESRF for
Claudio
Dear You Lin,
Siesta reads the pseudopotential in semilocal form and computes the
non-local
projectors (Kleiman-Bylander), so if you specify use.basis true then
siesta reads
from the files all the information it needs.
If you have just started using siesta I recommend the standard
Hi,
I attach an arch.make for the serial and parallel versions.
Both of them use the intel compiler and cmkl.
If you encounter any difficulties feel free to post them to the list.
Regards,
Eduardo
arch.make-intel-opteron
Description: Binary data
arch.make-intel-opteron-parallel
Description:
On nov 15, 2006, at 12:57 PM, SuiYang wrote:
Dear SuiYang,
¿Do you have the intel cluster math kernel library?
¿Do you have the intel mpi?
If not ¿is it possible that you install them?
If it is not possible to install them I will give you the detailed,
step by step,
instructions.
On nov 14, 2006, at 11:42 PM, Haiying He wrote:
Dear Haiying,
I don't have such a pseudo. My shortest convination
of pseudo/basis set for gold is:
%Block PAO.Basis
Au 3 0.23116
n=6 0 2 E15.16639 3.56453
4.26384 1.58867
1.0 1.0
n=6 1 1 E
Si en el fdf pones writexml (o un label equivalente, no me acuerdo)
funcionara
Edu
Jose:
Es algo que está ya en la lista de bug-fixes. Hasta que salga la
2.0.1 puedes arreglarlo a mano siguiendo las instrucciones que siguen.
Saludos,
Alberto
Inicio del mensaje
On oct 27, 2006, at 3:58 PM, Marcos Verissimo Alves wrote:
Hi Marcos,
What you are describing are semicore states: several shells with the
same l quantum number but different n quantum number.
All of them share the same pseudopotential, but each resulting
orbital is orthogonalized to those
On oct 24, 2006, at 2:53 PM, 박준호 wrote:
Hi,
The NODAT preprocessor variable was used to configure the precision
of several variables related to the paralelization.
In SIESTA 2.0 is deprecated and can be removed from the arch.make.
Regards,
Eduardo.
Hi, All
In source code, there is a
Hi Derek,
I haven't tried with lam-mpi, but with intel mpi there are problems
if you link with a blacs compiled for mpi 1.2 while using mpi 2
(lam-mpi implements almost all of mpi 2). The problem is in one
of the include files of the c interface of blacs, so it can be really
difficult to detect
Hi John,
I attach an arch.make for sgi-altix.
It uses the system blacs and scalapack optimized by sgi.
Regards,
Eduardo
arch.make-sgi
Description: Binary data
On sep 26, 2006, at 1:46 AM, John B. Baba wrote:
Hi all:
Thanks lan and Marcos replay,
I also try to compile siesta1.3
On sep 6, 2006, at 7:07 PM, Marcel Mohr wrote:
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
Hello all
I am trying to compile Siesta and required
Dear Sir,
There is a l_fce_pc_8.1* compiler. You can obtain a copy from your
premier account at intel.com
Regards,
Eduardo
On Wed, 2005-06-15 at 10:07 +0800, Mingsu Si wrote:
Dear Sir,
There is no l_fce_pc_8.1* compiler. In the web of Intel they say
l_fc_pc_8.1* support em64t CPU.
quotes
and it probably it will compile successfully.
Regards.-
simingsu wrote:
Dear Mr Eduardo Anglada
I have just started to compile the SIESTA code with Intel ifort 8.1
compiler and MKL 7.2.1 libraries. The problem is that it is not able to
compile atom.f file. The errors printed
You don't provide too much info, in this kind of reports the operating
system and cpu kind is really appreciated.
The compiler is saying it doesn't understand the symbols format of the
em64t lapack (mkl intel) library. Most probably you are compiling in
another architecture.
There are several
Dear Benjamin,
Siesta uses classical MD. In fact we have developed a new algorithm.
If you are interested the reference is:
http://dx.doi.org/10.1103/PhysRevE.68.055701
Efficient mixed-force first-principle molecular dynamics
E. Anglada, J. Junquera and J. M. Soler,
Phys. Rev. E *68*, 055701
Dear Imad,
The problem is in the 4 after Ga.
It should be a 3 as you only provide
3 shells. The fourth (corresponding to
the polarization orbital) will be added
automatically by siesta.
Regards,
Eduardo
%block PAO.Basis
Ga 3
n=3 2 2
0.00 0.00
n=4 0 2
1 - 100 of 104 matches
Mail list logo