[QE-users] Using symmetry in input structure description

[Husak Michal]

Thank you for the answer ...

To be 100% sure I had tested on diamond the computational time (all parameters set identical).


1) Material cloud generated file based on ibrav = 0 definition (BURAI generates the same)

scf time 22.11 sec


2) The same file but diamond was described by space group (space_group = 227 ! F d 3 m )

scf time 4.08 sec


So generally the usage of input files generated by Material Cloud and BURAI GUI should be avoided,

because they do not give the correct atom number and Bravais lattice and leed to longer computation times (not significant in my case, critical for more complex calculations).

The reason is sub optimal transfer from CIF files to the QE input not utilizing the information

about lattice centration ...


Michal