Are you using intel mkl library and gfortran for compiling?
I have faced this problem whenever i am using SOC and rel pps.
The solution is to do projwfc using a QE compiled version with netlib as the math library. You can still use the faster intel mkl for scf and nscf computations. Afterwards shift to qe compiled with netlib for projwfc.
Yasser Al Wahedi
Khalifa University
Assistant Professor
On Fri, Jan 19, 2018 at 7:00 PM +0400, "Akmal Khan"
<[email protected]> wrote:
file Mo.rel-pbe-spn-kjpaw_psl.0.2.UPF: wavefunction(s) 4D 4D renormalizedDear QE users and experts,I am doing SOC calcualtion with relativistic pseudopotentials using quantum espresso 5.4.0. But the projwfc.x for Pdos is exiting without any crash file. the output for pdos is shown below. Any idea about this problem will be very helpful.
file S.rel-pbe-n-kjpaw_psl.0.1.UPF: wavefunction(s) 3P 3P renormalized
file Si.rel-pbe-n-kjpaw_psl.0.1.UPF: wavefunction(s) 3S renormalized
file C.rel-pbe-n-kjpaw_psl.0.1.UPF: wavefunction(s) 2P 2P renormalized
Parallelization info
--------------------
sticks: dense smooth PW G-vecs: dense smooth PW
Min 47 47 15 6153 6153 1063
Max 48 48 16 6190 6190 1090
Sum 379 379 121 49325 49325 8601
Check: negative/imaginary core charge= -0.000016 0.000000
negative rho (up, down): 7.062E-03 0.000E+00
Gaussian broadening (read from input): ngauss,degauss= 1 0.009000
Calling projwave_nc ....
Thank You,
A.Khan
M.phil Research Student
Hazara University
Pakistan
_______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
