[Pw_forum] Parallel version and NFS

sir_puding at tut.by sir_puding at tut.by
Tue Apr 4 11:09:39 CEST 2006


Hi, All.

Does the calculations of SCF runs in parallel depend on type of data
storage? I'm using a Linux PIII cluster with 2 proc. per node and NFS
for data storage. and have long row of messages about unconverged
eigenvalues. In serial variant there is no such messages. I've noticed
that .wfcN files updated unsimultaneously. Or maybe i started
calculations in a wrong way?

here is my output

     Program PWSCF     v.3.1    starts ...
     Today is  4Apr2006 at 10:58:15

     Parallel version (MPI)

     Number of processors in use:      24
     K-points division:     npool     =    4
     R & G space division:  proc/pool =    6

     Ultrasoft (Vanderbilt) Pseudopotentials

     Current dimensions of program pwscf are:

     ntypx = 10   npk = 40000  lmax =  3
     nchix =  6  ndmx =  2000  nbrx = 14  nqfx =  8

     Starting configuration read from file     silicon.save

     Failed to open file     silicon.save

     Using input configuration

     looking for the optimal diagonalization algorithm ...

     dimension   time para (sec)   time serial (sec)
       300        1.11000000        3.49000000

     a parallel algorithm will be used for matrices larger than  300


     Planes per process (thick) : nr3 = 48 npp =   8 ncplane = 2304
 Proc/  planes cols    G   planes cols    G    columns  G
 Pool       (dense grid)      (smooth grid)   (wavefct grid)
  1      8    291   9118    8    291   9118   81   1351
  2      8    292   9118    8    292   9118   82   1350
  3      8    292   9118    8    292   9118   81   1349
  4      8    292   9117    8    292   9117   81   1349
  5      8    291   9117    8    291   9117   81   1349
  6      8    291   9117    8    291   9117   83   1349
  0     48   1749  54705   48   1749  54705  489   8097



     bravais-lattice index     =            0
     lattice parameter (a_0)   =      13.4834  a.u.
     unit-cell volume          =    2460.7699 (a.u.)^3
     number of atoms/cell      =           64
     number of atomic types    =            2
     kinetic-energy cutoff     =      30.0000  Ry
     charge density cutoff     =     120.0000  Ry
     convergence threshold     =      1.0E-08
     beta                      =       0.7000
     number of iterations used =            8  plain     mixing
     Exchange-correlation      =  SLA  PW   PBE  PBE (1434)
     nstep                     =           30

     celldm(1)=  13.483400  celldm(2)=   0.000000  celldm(3)=   0.000000
     celldm(4)=   0.000000  celldm(5)=   0.000000  celldm(6)=   0.000000

     crystal axes: (cart. coord. in units of a_0)
               a(1) = (  1.001285  0.002854  0.002847 )
               a(2) = (  0.002854  1.001309  0.002854 )
               a(3) = (  0.002847  0.002854  1.001285 )
     reciprocal axes: (cart. coord. in units 2 pi/a_0)
               b(1) = (  0.998733 -0.002839 -0.002831 )
               b(2) = ( -0.002839  0.998709 -0.002839 )
               b(3) = ( -0.002831 -0.002839  0.998733 )


     PSEUDO 1 is C  (US)    zval =  4.0   lmax= 2   lloc= 0
     Version   0  0  0 of US pseudo code
     Using log mesh of   627 points
     The pseudopotential has  4 beta functions with:
                l(1) =   0
                l(2) =   0
                l(3) =   1
                l(4) =   1
     Q(r) pseudized with  0 coefficients,  rinner =    0.000   0.000   0.000
                                                       0.000   0.000

     PSEUDO 2 is N  (US)    zval =  5.0   lmax= 2   lloc= 0
     Version   0  0  0 of US pseudo code
     Using log mesh of  1257 points
     The pseudopotential has  4 beta functions with:
                l(1) =   0
                l(2) =   0
                l(3) =   1
                l(4) =   1
     Q(r) pseudized with  0 coefficients,  rinner =    0.000   0.000   0.000
                                                       0.000   0.000

     atomic species   valence    mass     pseudopotential
        C              4.00    12.01078     C ( 1.00)
        N              5.00    14.00675     N ( 1.00)
      cell mass =  58.56510 AMU

     No symmetry!

   Cartesian axes

     site n.     atom                  positions (a_0 units)
         1           C   tau(  1) = (  -0.0004112  -0.0002048  -0.0004112  )
         2           C   tau(  2) = (   0.0020983   0.0022219   0.5009322  )
----
bla bla bla
----

     G cutoff =  552.6120  (  54705 G-vectors)     FFT grid: ( 48, 48, 48)

     nbndx  =  1200  nbnd   =   300  natomwfc =   256  npwx   =    1141
     nelec  = 257.00  nkb   =   512  ngl    =    9118
     
--------
bla bla bla
-------
     Cannot read rho : file not found

     Initial potential from superposition of free atoms

     starting charge  256.99668, renormalised to  257.00000
     Starting wfc from file

     total cpu time spent up to now is     12.84 secs

     Self-consistent Calculation

     iteration #  1     ecut=    30.00 ryd     beta=0.70
     Davidson diagonalization with overlap
     WARNING:    20 eigenvalues not converged
     WARNING:    18 eigenvalues not converged
     WARNING:    23 eigenvalues not converged
     WARNING:    33 eigenvalues not converged
     WARNING:    15 eigenvalues not converged
     WARNING:    19 eigenvalues not converged
     WARNING:    23 eigenvalues not converged
     WARNING:    19 eigenvalues not converged
     WARNING:    20 eigenvalues not converged
     WARNING:    26 eigenvalues not converged
     WARNING:    20 eigenvalues not converged
     WARNING:    25 eigenvalues not converged
     WARNING:    24 eigenvalues not converged
     WARNING:    21 eigenvalues not converged
     
Sergey

--
Please avoid sending me Word or PowerPoint attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html




More information about the users mailing list