My Community

Dear Sir,
I found that the calculating values for "Macroscopic dielectric properties" using VASP 5.4.4 will be different for using different cores in MPI no matter using the strategy via IBRION=6 or IBRION=8 accompanying with LCALCEPS=.TRUE. or LEPSILON=.TRUE.
But my calculations for optimization and band structures do not have this problem.
Could it be a bug in the code or it is a problem depending on how I compile VASP 5.4.4.

Best Regards,

I am not sure I understand your question.
Do you obtain different results for the dielectric tensor when you launch VASP in paralllel with a different number of cores?

Could you post the INCAR, OUTCAR, POSCAR, KPOINTS and POTCAR?