[QE-users] Out of memory in spin-polarized calculations?

Aulia Sukma Hutama aulia.sukma.hutama at ugm.ac.id
Sat Dec 24 11:53:14 CET 2022


Dear QE Developers and Users,

I am calculating electronic structure for some two-dimensional slab
materials. However, when I switch on the spin polarization calculations,
i.e., nspin = 2 and specify the total magnetization, the program crashes
(with obscured MPI errors) a few minutes after these lines

    Estimated total dynamical RAM >      13.41 GB

     Generating pointlists ...

     new r_m :   0.0687 (alat units)  1.6593 (a.u.) for type    1

     new r_m :   0.0744 (alat units)  1.7956 (a.u.) for type    2

     new r_m :   0.0687 (alat units)  1.6593 (a.u.) for type    3


     Check: negative core charge=   -0.000002


     Initial potential from superposition of free atoms


     starting charge  431.98888, renormalised to  432.00000

     Starting wfcs are  336 randomized atomic wfcs

     Checking if some PAW data can be deallocated...


     total cpu time spent up to now is      517.0 secs


     Self-consistent Calculation


     iteration #  1     ecut=    45.00 Ry     beta= 0.70

     Davidson diagonalization with overlap

I am running parallel using mpirun in 12 processor, using qe v7.1. I
suspected that the error comes due to insufficient available RAM since I
used to have 16 GB RAM in my desktop computer. However, I recently upgraded
my machine to a 32 GB RAM, but the problem persists. It only occurs when
spin polarized calculation is activated. I have scrolled through the
troubleshooting on the QE webpage without avail.

However, when I calculate using qe v.6.5 in the same machine, that
calculation works smoothly without any error and the results appear to be
consistent with the literature, even using my previous 16 GB machine. At
this moment, I will probably employ qe v6.5 for my project, but in the
future I will probably have to switch to qe v7.1 since it has more user
friendly input, especially for DFT+U(+V) calculations. I believe that I
compiled both versions using the same compiler, environment, etc. Can
anyone give insight into what happens in this case?. Please see below one
of the input files.

&CONTROL

  calculation = "scf",

  pseudo_dir  = "/home/kit/pseudo",

  prefix      = "TcS2"

/

&SYSTEM

  ibrav       = 0,

  nat         = 48,

  ntyp        = 3,

  ecutwfc     = 45.0,

  ecutrho     = 400.0,

  nbnd        = 300,

  nspin       = 2,

  tot_magnetization = 0.0,

/

&ELECTRONS

  conv_thr    = 1.D-6,

/

ATOMIC_SPECIES

Tc  1.0  Tc.pbe-spn-kjpaw_psl.0.3.0.UPF

S   1.0   S.pbe-n-kjpaw_psl.1.0.0.UPF

O   1.0   O.pbe-n-kjpaw_psl.1.0.0.UPF

CELL_PARAMETERS angstrom

       12.77686627          0.000439202         0.0000000000

        6.265655336         11.38463069         0.0000000000

        0.0000000000         0.0000000000        12.0000000000

ATOMIC_POSITIONS (angstrom)

Tc            4.8681648205        2.8282950764        3.5252508077

Tc            8.1008197205        8.6126403340        3.4465985039

Tc           11.3639244342        2.8784058563        3.4322457212

Tc           14.4814573750        8.5768900689        3.4611254097

Tc            6.1243824517        5.2843499195        3.4054551976

Tc            9.2859769258       11.1205084954        3.5079358365

Tc           12.5501517533        5.3022506153        3.4169441097

Tc           15.6741199532       11.0065008919        3.4286328276

Tc            7.7507445536        3.0270212606        3.4850520394

Tc           10.8988460162        8.7577382765        3.4795200336

Tc           14.2359961500        2.9788305239        3.5567234082

Tc           17.2968105508        8.7213588638        3.4983027213

Tc            3.3282446476        5.1443856525        3.4209441226

Tc            6.4739769360       10.8711632169        3.4131114537

Tc            9.7311834785        5.1660307518        3.3508718709

Tc           12.8666783326       10.8623579371        3.4045446585

O             3.1137920855        1.8340040263        2.8437314208

S             6.2571314096        7.4854751992        2.2543358478

S             9.5006356517        1.7928248869        2.2529842668

S            12.6411190583        7.4899016463        2.2587652221

S             7.9847223086        6.4038356725        4.5777392883

S             4.8962261249        0.6975177059        4.7296406164

S            14.3689559497        6.4084516267        4.6338911929

S            11.2379182648        0.7248293366        4.6234599177

S             6.3857249671        1.6210681936        2.1281352130

S             9.5843737492        7.3557709822        2.0824456746

S            12.9061272023        1.6249759542        2.1129453253

S            15.9711498518        7.3044166930        2.1167131928

S             4.6714178180        6.5610930509        4.7821288486

S             1.4695208609        0.8523433049        4.9063773474

S            11.0405869499        6.5714535109        4.7440592911

S             7.9263361365        0.8789461636        4.7849098351

S             4.7427209611        4.3507556506        1.7474840251

S             7.9816757552       10.2271554545        1.7587366509

S            11.2223705617        4.4335242802        1.6923051278

S            14.3312847875       10.1323762104        1.7207324077

S             6.2660255354        3.7405868577        5.1448279319

S             9.4350746439        9.4921695192        5.1486366434

S            12.7075366739        3.7421730369        5.1410100619

S            15.8245312495        9.4524306482        5.1701792036

S             7.8809512903        4.6588860216        1.8615919986

S            11.0189637439       10.4539443529        1.9127673915

S            14.3237148512        4.6134719981        1.9111670166

S            17.4334448548       10.3755910384        1.8930771966

S             3.1961933348        3.5770115187        5.1205949040

S             6.3440523273        9.2028784653        5.0032681734

S             9.5985204298        3.5271525709        4.9641169860

S            12.7275336090        9.2056351530        4.9996115212

K_POINTS automatic

 4 4 1 0 0 0


-- 
Aulia Sukma Hutama
Department of Chemistry
Universitas Gadjah Mada
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20221224/1907d87f/attachment.html>


More information about the users mailing list