My Community

I am trying to do slab calulations using 3 3 1 KPOINTS with PREC=MED which is taking longer time than we can afford.
Is it OK if i run 1 1 1 KPOINT calculation and then use the relaxed geometry for redoing the calculation using 3 3 1 KPOINTS.
Any suggestions or ways of handling this thing will be highly appreciated

thanx in advance

a Gamma-only calculation for a slab is rather risky (of course depending on the details of the electronic structure of your system). You can start some pre-convergence run with 1x1x1, but in any case you will have to check if the results (relaxed geometry) really corresponds to the equilibrium geometry by doing a few more ionic steps with an increased k-mesh.
another possibility to save computing time is to use soft PPs or to decrease the slab thickness (z). However, the vacuum size should be checked for convergence...