Re: [Pw_forum] compile attempt with external BLAS and intel MKL (espresso-5.2.1)

[Kane O'Donnell]

Hi Andrew,

This has happened before to me also, and it seems to happen because the login 
nodes for the XC40 have a slightly different environment than the compute 
nodes. The configure script is a little smarter on the MOM nodes it seems. I 
usually do the compilation in an interactive job on the debug (salloc -p debugq 
—nodes=1). Running a script on the MOM nodes is fine but make sure you use 
"aprun -n 1 -N 1 make" for the actual make step lest you upset your sysadmin. 
:o)

In the GNU environment, if you leave LAPACK_LIBS, SCALAPACK_LIBS, BLAS_LIBS, 
all that stuff (FFTs as well) blank and make sure they’re external, you will be 
fine. The Cray LibSci library is part of the “ftn” default linking environment 
and has all the correct symbols.

I agree, the intel environment is a bit harder if you want MKL. You could just 
use the intel environment with the Cray libsci if you want the extra 
optimisations from ifort. Both MKL and Cray use an FFTW implementation anyway 
for Quantum Espresso anyway don’t they? No real benefit to MKL unless you’re 
after 10%-level optimisations which seems like a waste of time unless you’re 
actually one of the HPC technicians who run the machine.

Best regards,

Kane


> On 11 Nov 2015, at 21:48, Andrew Downs  wrote:
> 
> Kane,
>  
> Thank you, your suggestions worked, although I think I had to go a step 
> further and get into make.sys and comment out the line where it found the 
> intel BLAS libraries and point it to the internal BLAS.
> Something about the intel MKL trips up compilation for us, my other 
> researcher isn't able to get it to work either and he's got a good 20 years 
> of experience on me.
>  
> -Andrew
>  
> Sent: Monday, November 09, 2015 at 10:23 PM
> From: "Kane O'Donnell" 
> To: "PWSCF Forum" 
> Subject: Re: [Pw_forum] compile attempt with external BLAS and intel MKL 
> (espresso-5.2.1)
>  
> Hi Andrew,
>  
> Are you trying to just build espresso 5.2.1, or are you trying to build it 
> using some custom linear algebra libraries? On a Cray you’re encouraged 
> strongly *not* to stuff around with custom BLAS/FFT/scalapack/etc. The 
> programming environments are set up so you don’t need to do that, and for 
> example on our XC40 you can build SVN espresso just with:
>  
> module swap PrgEnv-cray PrgEnv-gnu
> ./configure —enable-parallel —with-scalapack=yes FC=ftn CC=cc 
> make all
>  
> You can do the same for PrgEnv-intel (it’s exactly the same configure line) 
> but it’s still wise to use the MKL link line advisor to change make.sys as 
> necessary - you can copy and paste directly from the link line advisor as 
> Cray set $MKLROOT properly when you load PrgEnv-intel.
>  
> Our cluster guys made a wiki entry about QE on the XC40 using the intel suite 
> here:
>  
> https://portal.ivec.org/docs/Supercomputers/Magnus/Quantum_Espresso 
> 
>  
> It’s not quite the same as the link line advisor but it works (except 
> XSpectra, you need the gnu environment to get that to work on our XC40 for 
> some reason). 
>  
> Kane
>  
>  
> On 10 Nov 2015, at 01:10, Andrew Downs  > wrote:
>  
> Hello,
>  
> I've been able to jump a few hurdles that have come up in my compile process, 
> but this one has me stumped.
> I'm compiling on a Cray XC-40, and another researcher here was able to build 
> 5.1.1 with out BLAS and MKL libraries.
> He and I assumed 5.2.1 should also work under the same circumstances, but 
> this may turn out not to be the case.
>  
> my configure line is as follows:
> 
> ./configure ARCH=x86_64 MPIF90=ftn FFLAGS="-O3" CC=icc CFLAGS=-O3 
> --prefix=/usr/cta/unsupported/qe/espresso-5.2.1
>  
> relevant part of make.sys:
>  
> LDFLAGS=--static 
> -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/  
> -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/intel64/lp64/
> BLAS_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_sequential.a
>  -Wl,--end-group 
> /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.a
>  -lpthread -lm
> LAPACK_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_intel_lp64.a
>  
> /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_core.a
> SCALAPACK_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_scalapack_lp64.a
>  -Wl,--start-group
> FFT_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_cdft_core.a
>  
>  
> make all output:
> ftn --static 
> -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/  
> -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/intel64/lp64/
>  -o pw.x \
>pwscf.o  libpw.a ../../Modules/libqemod.a ../../flib/ptools.a 
> ../../flib/flib.a ../../clib/clib.a ../../iotk/src/libiotk.a 
> /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_scalapack_lp64.a
>  -Wl,-

[Pw_forum] GIPAW Problem (Cholesky)

[Vic Bermudez]

Hello,

I'm trying to use GIPAW for the first time and encountering a problem 
that
isn't discussed at the users' forum. I'm using GIPAW vers. v.5.0.2 (svn rev.
9392), and I get this error message soon after execution starts:


 Error in routine  cdiaghg (2785):
  problems computing cholesky
 
 stopping ...

Here's my GIPAW input file:

&INPUTGIPAW
job='nmr',
tmp_dir='/lustre/cmf/scratch/b/bermudez/XX_166928/',
iverbosity=1,
restart_mode='from_scratch',
max_seconds=259000.0,
use_nmr_macroscopic_shape=.false.
/
***

And here's the GIPAW output right up to the point where 
the error message
occurs:
***
Parallel version (MPI), running on64 processors
 R & G space division:  proc/nbgrp/npool/nimage =  64

   Info: using nr1, nr2, nr3 values from input

   Info: using nr1s, nr2s, nr3s values from input

 IMPORTANT: XC functional enforced from input :
 Exchange-correlation  =  SLA  PW   PBX  PBC ( 1 4 3 4 0)
 EXX-fraction  =0.00
 Any further DFT definition will be discarded
 Please, verify this is what you really want

   file Zr.pbe-spn-kjpaw_psl.0.2.3.UPF: wavefunction(s)  4D
renormalized
   file O.pbe-n-kjpaw_psl.0.1.UPF: wavefunction(s)  2P
renormalized
   file H.pbe-kjpaw_psl.0.1.UPF: wavefunction(s)  1S
renormalized

 Parallelization info
 
 sticks:   dense  smooth PW G-vecs:dense   smooth  PW
 Min 382 153 38   138228349494368
 Max 383 154 39   138247350064379
 Sum   244919807   2451  8847271  2238233  279837


 Check: negative/imaginary core charge=   -0.010.00

 negative rho (up, down):  0.128E-01 0.000E+00

 Subspace diagonalization in iterative solution of the eigenvalue
problem:
 scalapack distributed-memory algorithm (size of sub-group:  5*  5
procs)

 init_paw_1: ntyp= 1  rc=1.6000  rs=1.0667
 init_paw_1: ntyp= 1  rc=1.6000  rs=1.0667
 init_paw_1: ntyp= 1  rc=1.7000  rs=1.1333
 init_paw_1: ntyp= 1  rc=1.7000  rs=1.1333
 init_paw_1: ntyp= 1  rc=1.9000  rs=1.2667
 init_paw_1: ntyp= 1  rc=1.9000  rs=1.2667

 init_gipaw_1: projectors nearly linearly dependent:
 ntyp =  1, l/n1/n2 =  2 2 1  0.99876687
 init_paw_1: ntyp= 2  rc=1.3500  rs=0.9000
 init_paw_1: ntyp= 2  rc=1.3500  rs=0.9000
 init_paw_1: ntyp= 2  rc=1.3500  rs=0.9000
 init_paw_1: ntyp= 2  rc=1.3500  rs=0.9000
 init_gipaw_1: projectors nearly linearly dependent:
 ntyp =  2, l/n1/n2 =  0 2 1  0.99100135
 init_gipaw_1: projectors nearly linearly dependent:
 ntyp =  2, l/n1/n2 =  1 2 1  0.99790100
 init_paw_1: ntyp= 3  rc=1.  rs=0.6667
 init_paw_1: ntyp= 3  rc=1.  rs=0.6667
 init_gipaw_1: projectors nearly linearly dependent:
 ntyp =  3, l/n1/n2 =  0 2 1  0.99968852

 Message from routine gipaw_setup:
 * implemented only for insulators *

 GIPAW job: nmr
 NMR macroscopic correction: no

 GIPAW:  0m54.34s CPU 1m 1.81s WALL

 Computing the magnetic susceptibility isolve=0ethr=0.1000E-13
 Starting from scratch
 k-point #1 of 1  pool #  1
***

I notice that there are statements like "projectors nearly linearly
dependent". I'm not sure what this means, but it can't be good. I should
also note that this calculation is being done for a rather large system
(1280 electrons). I would greatly appreciate any guidance. Thank you in
advance.

Best Wishes,
Vic Bermudez

Victor M. Bermudez
E-mail: bermu...@alum.mit.edu


___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Dielectric constant

[Giuseppe Mattioli]

Dear Sohail Ahmad
The error message is quite clear: epsilon.x works only with
norm-conserving pseudopotentials (NCPP), while you are using
Vanderbilt's ultrasoft pseudopotentials (USPP).
HTH
Giuseppe

Giuseppe Mattioli
ISM-CNR
Italy

Quoting Sohail Ahmad :

> Dear QE users,
>
> I wish to calculate dielectric constants and some other related properties.
> I am getting following errors. My input file (scf and eps) are attached.
> I am using QE-5.0.1
> 
>  Fermi energy [eV] is: 12.49326
>  The system is a metal...
>
>  Performing eps calculation...
>
>   
> %%
>  Error in routine grid_build (1):
>  USPP are not implemented
>   
> %%
>
> &inputpp
>  prefix = 'ScN',
>  outdir = './OUT',
>  calculation = 'eps',
> /
> &ENERGY_GRID
>  smeartype = 'gaussian',
>  intersmear = 0.1,
>  wmin = 0.00
>  wmax = 10.00
>  nw = 5000
> /
> 
> &control
>   calculation = 'scf',
>   restart_mode = 'from_scratch',
>   prefix = 'ScN',
>   pseudo_dir = '/home/sohail/Espresso/espresso-5.0.1/pseudo',
>   outdir = './OUT',
>  /
>  &system
>   ibrav = 2, a = 4.501,
>   nat = 2, ntyp = 2,
>   ecutwfc = 250,
>   ecutrho = 1000,
>   nbnd = 20,
>   occupations = 'smearing', smearing = 'gaussian', degauss = 0.001,
>   nspin = 2,
>   starting_magnetization(1) = 0.1d0,
>  /
>  &electrons
>   mixing_beta = 0.3,
>   conv_thr = 1.0d-9,
>  /
> ATOMIC_SPECIES
> Sc  44.956  Sc.pw91-nsp-van.UPF
> N   14.007  N.pw91-van_ak.UPF
> ATOMIC_POSITIONS {angstrom}
> Sc  0.00  0.00  0.00
> N   2.250500  2.250500  2.250500
> K_POINTS AUTOMATIC
> 12 12 12 1 1 1
>
> -
> SOHAIL AHMAD
> KING KHALID UNIVERSITY
> SAUDI ARABIA
> ___
> Pw_forum mailing list
> Pw_forum@pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum


-- 

- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.


Giuseppe Mattioli
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
v. Salaria Km 29,300 - C.P. 10
I 00015 - Monterotondo Stazione (RM)
Tel + 39 06 90672836 - Fax +39 06 90672316
E-mail: 

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[Giuseppe Mattioli]

Dear Phil

> But I mentioned a problem with using a very dense k - point grid and an 8
> atom Si cell, the program exits before it can output the forces, stresses,
> and the individual contributions to the total energy

Sorry, but this was not mentioned in the current thread in which you  
asked how to calculate the band gap...

> Nobody was able to point me in the right direction with that. Also, another
> user has previously reported that using a 2 atom si cell results in an
> oscillation in the total energy as you increase ecutwfc (something I have
> also verified)

Have you checked if the Gamma gap or the "best" gap (at fixed  
occupations) jumps like the total energy or if it is already converged?

However, both Stefano's and Mostafa's suggestions deserve to be considered.

Have a nice computation and don't forget to add your affiliation to  
the posts to this forum.

Giuseppe

Giuseppe Mattioli
ISM-CNR
Italy

Quoting Ref Fymz :

> But I mentioned a problem with using a very dense k - point grid and an 8
> atom Si cell, the program exits before it can output the forces, stresses,
> and the individual contributions to the total energy (hence why regardless
> of what screening parameter I used I got the same band gap, as the fock).
> Nobody was able to point me in the right direction with that. Also, another
> user has previously reported that using a 2 atom si cell results in an
> oscillation in the total energy as you increase ecutwfc (something I have
> also verified), so I'm not sure I can accurately calculate the band gap
> using a 2 atom cell either.
>
> Phil
>
> On 11 November 2015 at 12:18, Ref Fymz  wrote:
>
>> Hey,
>>
>> Thanks for the reply. So if I use an 8 atom Si cell, and use fixed
>> occupations with 32 bands, the difference between the highest occupied and
>> lowest unoccupied is my actual band gap? (i.e the one I can compare to PBE
>> and experiment)
>>
>>
>> Phil
>>
>> On 11 November 2015 at 10:13, Ref Fymz  wrote:
>>
>>> Hey,
>>>
>>> I've tried searching the forum, and I haven't come across a way to
>>> calculate the band gap in a material using the hybrid functionals. I know
>>> you can't do nscf calculations, but is there any simple way to calculate
>>> the band gap for a material using scf simulations?
>>>
>>> Thanks,
>>>
>>> Phil
>>>
>>
>>


-- 

- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.


Giuseppe Mattioli
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
v. Salaria Km 29,300 - C.P. 10
I 00015 - Monterotondo Stazione (RM)
Tel + 39 06 90672836 - Fax +39 06 90672316
E-mail: 

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[stefano de gironcoli]

Dear Giuseppe, dear all,

   I think that it should be possible to
   -1) run an scf hybrid calculation with a number of empty bands
   -2) use wannier90 to extract the MLWF (meaning the shortest-ranged TB 
hamiltonian that gives those eigenvalues). Some dependence on the 
disentanglement of the conduction bands would result but probably it 
would not affect low-lying CB states.
   -3) draw the k-dependent band structure of the resulting TB model 
that should reproduce the full calculation in the points that where 
present in the original scf calculation.

   for dense enough grids everything should make sense and converge to 
the right results.
   how fast this happens I don't know but I would be hopeful

   good luck

stefano

On 11/11/2015 14:09, Giuseppe Mattioli wrote:
> Dear all
>
>> So if I use an 8 atom Si cell, and use fixed
>> occupations with 32 bands, the difference between the highest occupied and
>> lowest unoccupied is my actual band gap?
> This is true for direct band gap semiconductors, where both VBM and
> CBM can be found at Gamma, which must be included in the list. This is
> easy because every N, N, N, 0, 0, 0 automatically generated grid
> includes Gamma. However, in the case of bulk Si the CBM falls
> somewhere along the Gamma-X direction, close to X. If your k-point
> mesh is *very* dense, then the CBM found by the code at a given
> k-point will be negligibly shifted wrt the proper CBM. Otherwise you
> should perform the calculation by using the proper 2-atom Si cell
> (faster calculation with EXX functionals...) and you should be sure to
> include one k-point close to the CBM in the mesh. You can still work
> with the sc 8-atom cell, if you prefer. But you must *fold* the 8-atom
> sc cell into the 2-atom fcc cell.
> HTH
> Giuseppe
>
> Giuseppe Mattioli
> ISM-CNR
> Italy
>
> Quoting Ref Fymz :
>
>> Hey,
>>
>> Thanks for the reply. So if I use an 8 atom Si cell, and use fixed
>> occupations with 32 bands, the difference between the highest occupied and
>> lowest unoccupied is my actual band gap? (i.e the one I can compare to PBE
>> and experiment)
>>
>>
>> Phil
>>
>> On 11 November 2015 at 10:13, Ref Fymz  wrote:
>>
>>> Hey,
>>>
>>> I've tried searching the forum, and I haven't come across a way to
>>> calculate the band gap in a material using the hybrid functionals. I know
>>> you can't do nscf calculations, but is there any simple way to calculate
>>> the band gap for a material using scf simulations?
>>>
>>> Thanks,
>>>
>>> Phil
>>>
>

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

[Pw_forum] Dielectric constant

[Sohail Ahmad]

Dear QE users,

I wish to calculate dielectric constants and some other related properties.
I am getting following errors. My input file (scf and eps) are attached.
I am using QE-5.0.1

 Fermi energy [eV] is: 12.49326
 The system is a metal...

 Performing eps calculation...

 %%
 Error in routine grid_build (1):
 USPP are not implemented
 %%

&inputpp
 prefix = 'ScN',
 outdir = './OUT',
 calculation = 'eps',
/
&ENERGY_GRID
 smeartype = 'gaussian',
 intersmear = 0.1,
 wmin = 0.00
 wmax = 10.00
 nw = 5000
/

&control
  calculation = 'scf',
  restart_mode = 'from_scratch',
  prefix = 'ScN',
  pseudo_dir = '/home/sohail/Espresso/espresso-5.0.1/pseudo',
  outdir = './OUT',
 /
 &system
  ibrav = 2, a = 4.501, 
  nat = 2, ntyp = 2,
  ecutwfc = 250,
  ecutrho = 1000,
  nbnd = 20,
  occupations = 'smearing', smearing = 'gaussian', degauss = 0.001,
  nspin = 2,
  starting_magnetization(1) = 0.1d0,
 /
 &electrons
  mixing_beta = 0.3,
  conv_thr = 1.0d-9,
 /
ATOMIC_SPECIES
Sc  44.956  Sc.pw91-nsp-van.UPF
N   14.007  N.pw91-van_ak.UPF
ATOMIC_POSITIONS {angstrom}
Sc  0.00  0.00  0.00
N   2.250500  2.250500  2.250500
K_POINTS AUTOMATIC
12 12 12 1 1 1

- 
SOHAIL AHMAD
KING KHALID UNIVERSITY
SAUDI ARABIA
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[Mostafa Youssef]

Dear Phil and all,


Note that for a functional with exact exchange, the difference between 
ionization potential and electron affinity (I-A) is not the same as the 
Kohn-Sham single particle states  gap.  The consensus, as far as I know, is 
that I-A is more physical and relevant to comparison with experimental band 
gaps.  Calculating I-A  requires 3 scf calculations (neutral cell, +electorn, 
-electron).  The cases of +/- e require high smearing similar to that of metals 
(or you could use the tetrahedron method but I doubt it is implemented for 
hybrids). Also +/-  calculations require large cells and correction for a 
charged cell (Makov-Payne for example).  See section III-A in this paper for 
elaboration on computational details:

PHYSICAL REVIEW B 78, 235104 ?2008?

For PBE, LDA, ...  I-A is the same as  Kohn-Sham gap.

For elaboration on the theoretical aspects and why I-A is not the same as K-S 
band gap for the exact exchange, consult the work of A. J. Cohen and 
collaborators .

Mostafa
MIT
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[Ref Fymz]

But I mentioned a problem with using a very dense k - point grid and an 8
atom Si cell, the program exits before it can output the forces, stresses,
and the individual contributions to the total energy (hence why regardless
of what screening parameter I used I got the same band gap, as the fock).
Nobody was able to point me in the right direction with that. Also, another
user has previously reported that using a 2 atom si cell results in an
oscillation in the total energy as you increase ecutwfc (something I have
also verified), so I'm not sure I can accurately calculate the band gap
using a 2 atom cell either.

Phil

On 11 November 2015 at 12:18, Ref Fymz  wrote:

> Hey,
>
> Thanks for the reply. So if I use an 8 atom Si cell, and use fixed
> occupations with 32 bands, the difference between the highest occupied and
> lowest unoccupied is my actual band gap? (i.e the one I can compare to PBE
> and experiment)
>
>
> Phil
>
> On 11 November 2015 at 10:13, Ref Fymz  wrote:
>
>> Hey,
>>
>> I've tried searching the forum, and I haven't come across a way to
>> calculate the band gap in a material using the hybrid functionals. I know
>> you can't do nscf calculations, but is there any simple way to calculate
>> the band gap for a material using scf simulations?
>>
>> Thanks,
>>
>> Phil
>>
>
>
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] compile attempt with external BLAS and intel MKL (espresso-5.2.1)

[Andrew Downs]

Kane,

 

Thank you, your suggestions worked, although I think I had to go a step further and get into make.sys and comment out the line where it found the intel BLAS libraries and point it to the internal BLAS.

Something about the intel MKL trips up compilation for us, my other researcher isn't able to get it to work either and he's got a good 20 years of experience on me.

 

-Andrew

 

Sent: Monday, November 09, 2015 at 10:23 PM
From: "Kane O'Donnell" 
To: "PWSCF Forum" 
Subject: Re: [Pw_forum] compile attempt with external BLAS and intel MKL (espresso-5.2.1)



 
Hi Andrew,

 


Are you trying to just build espresso 5.2.1, or are you trying to build it using some custom linear algebra libraries? On a Cray you’re encouraged strongly *not* to stuff around with custom BLAS/FFT/scalapack/etc. The programming environments are set up so you don’t need to do that, and for example on our XC40 you can build SVN espresso just with:

 

module swap PrgEnv-cray PrgEnv-gnu

./configure —enable-parallel —with-scalapack=yes FC=ftn CC=cc 

make all

 

You can do the same for PrgEnv-intel (it’s exactly the same configure line) but it’s still wise to use the MKL link line advisor to change make.sys as necessary - you can copy and paste directly from the link line advisor as Cray set $MKLROOT properly when you load PrgEnv-intel.

 

Our cluster guys made a wiki entry about QE on the XC40 using the intel suite here:

 

https://portal.ivec.org/docs/Supercomputers/Magnus/Quantum_Espresso

 

It’s not quite the same as the link line advisor but it works (except XSpectra, you need the gnu environment to get that to work on our XC40 for some reason). 

 

Kane

 

 




On 10 Nov 2015, at 01:10, Andrew Downs  wrote:
 




Hello,

 

I've been able to jump a few hurdles that have come up in my compile process, but this one has me stumped.

I'm compiling on a Cray XC-40, and another researcher here was able to build 5.1.1 with out BLAS and MKL libraries.

He and I assumed 5.2.1 should also work under the same circumstances, but this may turn out not to be the case.

 

my configure line is as follows:


./configure ARCH=x86_64 MPIF90=ftn FFLAGS="-O3" CC=icc CFLAGS=-O3 --prefix=/usr/cta/unsupported/qe/espresso-5.2.1

 

relevant part of make.sys:

 


LDFLAGS=--static -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/  -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/intel64/lp64/
BLAS_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_sequential.a -Wl,--end-group /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.a -lpthread -lm
LAPACK_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_intel_lp64.a /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_core.a
SCALAPACK_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_scalapack_lp64.a -Wl,--start-group
FFT_LIBS=/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_cdft_core.a

 


 

make all output:

ftn --static -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/  -I/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/include/intel64/lp64/ -o pw.x \
   pwscf.o  libpw.a ../../Modules/libqemod.a ../../flib/ptools.a ../../flib/flib.a ../../clib/clib.a ../../iotk/src/libiotk.a /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_scalapack_lp64.a -Wl,--start-group /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_intel_lp64.a /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_core.a /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_cdft_core.a /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_sequential.a -Wl,--end-group /opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.a -lpthread -lm
/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_core.a(mkl_memory.o): In function `mkl_serv_set_memory_limit':
../../../../serv/kernel/mkl_memory.c:(.text+0x521): warning: Using 'dlopen' in statically linked applications requires at runtime the shared libraries from the glibc version used for linking
/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.a(dgesd2d_.o): In function `dgesd2d_':
../../../../scalapack/BLACS/SRC/MPI/dgesd2d_.c:(.text+0x92): undefined reference to `mkl_serv_get_mpi_wrappers'
/opt/intel/compilers_and_libraries_2016.0.109/linux/mkl/lib/intel64/libmkl_blacs_intelmpi_lp64.a(dgesd2d_.o): In function `dsendrecv_':
../../../../scalapack/BLACS/SRC/MPI/dgesd2d_.c:(.text+0x142): undefined reference to `mkl_serv_get_mpi_wrappers'
../../../../scalapack/BLACS/SRC/MPI/dgesd2d_.c:(.text+0x162): undefined reference to `mkl_serv_get_mpi_wrappers'
../../../../scalapack/BLACS/SRC/MPI/dgesd2d_.c:(.text+0x1aa)

Re: [Pw_forum] Band gap hybrid functionals

[Giuseppe Mattioli]

Dear all

> So if I use an 8 atom Si cell, and use fixed
> occupations with 32 bands, the difference between the highest occupied and
> lowest unoccupied is my actual band gap?

This is true for direct band gap semiconductors, where both VBM and  
CBM can be found at Gamma, which must be included in the list. This is  
easy because every N, N, N, 0, 0, 0 automatically generated grid  
includes Gamma. However, in the case of bulk Si the CBM falls  
somewhere along the Gamma-X direction, close to X. If your k-point  
mesh is *very* dense, then the CBM found by the code at a given  
k-point will be negligibly shifted wrt the proper CBM. Otherwise you  
should perform the calculation by using the proper 2-atom Si cell  
(faster calculation with EXX functionals...) and you should be sure to  
include one k-point close to the CBM in the mesh. You can still work  
with the sc 8-atom cell, if you prefer. But you must *fold* the 8-atom  
sc cell into the 2-atom fcc cell.
HTH
Giuseppe

Giuseppe Mattioli
ISM-CNR
Italy

Quoting Ref Fymz :

> Hey,
>
> Thanks for the reply. So if I use an 8 atom Si cell, and use fixed
> occupations with 32 bands, the difference between the highest occupied and
> lowest unoccupied is my actual band gap? (i.e the one I can compare to PBE
> and experiment)
>
>
> Phil
>
> On 11 November 2015 at 10:13, Ref Fymz  wrote:
>
>> Hey,
>>
>> I've tried searching the forum, and I haven't come across a way to
>> calculate the band gap in a material using the hybrid functionals. I know
>> you can't do nscf calculations, but is there any simple way to calculate
>> the band gap for a material using scf simulations?
>>
>> Thanks,
>>
>> Phil
>>


-- 

- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.


Giuseppe Mattioli
CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
v. Salaria Km 29,300 - C.P. 10
I 00015 - Monterotondo Stazione (RM)
Tel + 39 06 90672836 - Fax +39 06 90672316
E-mail: 

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[Ref Fymz]

Hey,

Thanks for the reply. So if I use an 8 atom Si cell, and use fixed
occupations with 32 bands, the difference between the highest occupied and
lowest unoccupied is my actual band gap? (i.e the one I can compare to PBE
and experiment)


Phil

On 11 November 2015 at 10:13, Ref Fymz  wrote:

> Hey,
>
> I've tried searching the forum, and I haven't come across a way to
> calculate the band gap in a material using the hybrid functionals. I know
> you can't do nscf calculations, but is there any simple way to calculate
> the band gap for a material using scf simulations?
>
> Thanks,
>
> Phil
>
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[DHIRENDRA VAIDYA]

vasp like method was giving some  n+q error. DOn't know if that has been
fixed in 5.2.1. But to calculate bandgap you can add enough conduction
bands and use occupations='fixed' that gives you highest occupied and
lowest unoccupied level in the out file.

On Wed, Nov 11, 2015 at 4:43 PM, plgong  wrote:

> you can obtained band structure using hse method, but
> first you must modify your K_points in your inputfile  like
> the method using vasp!
>
> 在2015-11-11 18:13:41，plgongplg...@theory.issp.ac.cn写道：
>
> Hey,
>
> I've tried searching the forum, and I haven't come across a way to
> calculate the band gap in a material using the hybrid functionals. I know
> you can't do nscf calculations, but is there any simple way to calculate
> the band gap for a material using scf simulations?
>
> Thanks,
>
> Phil
>
>
>
>
>
> ___
> Pw_forum mailing list
> Pw_forum@pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
--
Dhirendra
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] Band gap hybrid functionals

[plgong]

you can obtained band structure using hse method, but
first you must modify your K_points in your inputfile  like
the method using vasp!

在2015-11-11 18:13:41，plgongplg...@theory.issp.ac.cn写道：

Hey,


I've tried searching the forum, and I haven't come across a way to calculate 
the band gap in a material using the hybrid functionals. I know you can't do 
nscf calculations, but is there any simple way to calculate the band gap for a 
material using scf simulations?


Thanks,


Phil


___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

[Pw_forum] Band gap hybrid functionals

[Ref Fymz]

Hey,

I've tried searching the forum, and I haven't come across a way to
calculate the band gap in a material using the hybrid functionals. I know
you can't do nscf calculations, but is there any simple way to calculate
the band gap for a material using scf simulations?

Thanks,

Phil
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

[Pw_forum] need help about incorporation energy calculation

[Shaofeng Wang]

Dear pwscf experts,

I am doing some calculations about incorporation of arsenate ions into 
lattice of gypsum. But I am not clear how to calculate the overall reaction 
energy. For example, I calculated the total energy of gypsum supercell 
(2*1*2) and arsenate incorporated gypsum supercell. But for AsO43- and SO42- 
ions, I just calculated the total energy of AsO4 with -3 charge and SO4 
group with -2 charge in a cubic box (15 * 15 * 15A). Then the overall 
reaction energy was calculated from the difference between the total energy 
of products and reactants as following equation.
Ca16S16O96H64 + AsO43- = Ca16AsS15O96H64- + SO42-

Did I do right thing?

If not, can any expert show me how to calculate correctly?

By the way, one more question is the lower total energy means more stable 
crystal?

Thanks in advance.

Best,

Shaofeng


--
Shaofeng Wang, Ph.D of Geochemistry
Environmental Molecular Science Group
Institute of Applied Ecology, Chinese Academy of Sciences
Shenyang, 110016, China
wangshaof...@iae.ac.cn
www.iae.cas.cn


___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

[Pw_forum] pp.x input

[Jaret Qi]

Dear QE users,When I try to visualize spin polarization and choose iflag = 0 ( 
1D plot of the spherical average), and output_format = 0 (format suitable for 
gnuplot), then I've got an output with three columns. but I do not know what 
does each column represent?Also when I use iflag=1 and output_format = 0, I got 
2 columns, again what does they represent?
any help?
-jaretASU___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

users@lists.quantum-espresso.org

[Jaret Qi]

Hi all,when I run a vc-relax with spin true using version 5.2.0, it shows the 
magnetization of each atom as well as the total and absolute magnetization 
during relaxation run. But using 5.0.2 version, it shows only total and 
absolute magnetization. My concern is how to get the magnetization of each atom 
using 5.0.2 version?I guess I should use pdos, am I right? 
best,jaretASU___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum