<html><body><div style="font-family: arial, helvetica, sans-serif; font-size: 12pt; color: #000000"><div><span style="font-family: monospace;" data-mce-style="font-family: monospace;"><span style="color: #000000; background-color: #ffffff;" data-mce-style="color: #000000; background-color: #ffffff;">Hi dear QE users, </span><br><br>I am trying to capture the band structure of PbSe - a semiconductor with a small gap. <br>The literature says that this is a tricky case requiring a hybrid functional treatment <br>(Hummer et al PhysRevB 75, 195211 (2007)). <br>I did such a calculation with QE-7.3.1 compiled locally on a 6-core machine. <br>The calculation ends without problems. </span></div><div><span style="font-family: monospace;" data-mce-style="font-family: monospace;"><br>I decided to try to use a bigger HPC cluster with QE-7.2.2 <br>and as a test I ran the very same calculation which now fails <br>with the infamous "charge error" message. </span></div><div><span style="font-family: monospace;" data-mce-style="font-family: monospace;">This happens with the same number of MPI threads and also with more threads.<br><br>There are not so many pseudos recommended for HSE so I use SG15 variants <br>in both cases. I just perform the SCF run of pw.x because the K points <br>include the L point where the gap is. I don't need the high-res non SCF <br>calculation (yet) which is anyway not possible with HSE. <br><br>the input file : <br><br>======================================================================= <br>&CONTROL <br> outdir = 'scratch/' <br> prefix = 'PbSe' <br> pseudo_dir = '/home/tjoli/handson_pwscf/sg15pseudos/' <br> calculation='scf' <br> nstep = 200 <br> wf_collect = .true. <br> verbosity='high' <br>/ <br><br>&SYSTEM <br> celldm(1) = 11.6 <br> ecutwfc = 48.0 <br> ecutrho = 300.0 <br> nbnd = 24 <br> ibrav = 2 <br> lspinorb = .true. <br> nat = 2 <br> noncolin = .true. <br> ntyp = 2 <br> input_dft='hse' <br> nqx1 = 3 <br> nqx2 = 3 <br> nqx3 = 3 <br>/ <br><br>&ELECTRONS <br> conv_thr = 1e-8 <br> diagonalization = "cg" <br> mixing_beta = 0.5 <br> mixing_mode = 'plain' <br> startingwfc = 'atomic-random' <br>/ <br><br>ATOMIC_SPECIES <br> Pb 207.2 Pb_ONCV_PBE_FR-1.0.upf <br> Se 78.96 Se_ONCV_PBE_FR-1.1.upf <br><br><br>ATOMIC_POSITIONS alat <br> Pb 0.000000000000 0.000000000000 0.000000000000 <br> Se 0.500000000000 0.000000000000 0.000000000000 <br><br>K_POINTS automatic <br> 6 6 6 0 0 0 <br><br>==================================================================== <br>the error message: <br><br> ethr = 1.00E-02, avg # of iterations = 10.9 <br><br> negative rho (up, down): 1.120E+00 0.000E+00 <br><br> total cpu time spent up to now is 2.4 secs <br><br> WARNING: integrated charge= 10.23051228, expected= 20.00000000 <br><br>%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% <br> Error in routine electrons (1): <br> charge is wrong <br>%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% <br><br> stopping ... <br><br>====================================================================== <br><br>the same calculation on my local QE-7.3.1 gives : <br><br> starting charge 19.8764, renormalised to 20.0000 <br><br>which is what is expected from the electron numbers in the pseudos. <br><br><br>hints appreciated !<br>Th Jolicoeur<br></span></div><div><span style="font-family: monospace;" data-mce-style="font-family: monospace;"><br data-mce-bogus="1"></span></div><div><span style="font-family: monospace;" data-mce-style="font-family: monospace;">senior researcher CNRS</span></div><div data-marker="__SIG_PRE__">==================================================================<br>Thierry Jolicoeur<br>Institut de Physique Theorique-DRF<br>room 16 blg 774<br>Orme des Merisiers<br>91190 Gif sur Yvette<br>tel +33 169087512<br>P44W+V7 Saint-Aubin<br>ICBM: 48.708004N, 2.145753E<br><br>"It is unworthy of excellent men to lose hours like slaves <br>in the labour of calculation which could safely be <br>relegated to anyone else if machines were used." <br>Gottfried Wilhelm Leibniz<br>==================================================================</div></div></body></html>