Density functional theory (DFT) has become an extremely popular method of solving electronic structure problems in chemistry and solid-state physics.
In solid-state physics DFT results and predictions become more accurate year by year.
But diminishing returns have started to appear in the accuracy of the popular generalized gradient approximation (GGA) level of DFT theory.
In order to overcome this fundamental limitation of GGA level we must either move on to more sophisticated levels of theory, such as meta-GGAs or exact-exchange hybrid functionals,
or explore ways of somehow extending the limits of GGA level.
In this talk I focus on the latter option and introduce one way of extending the accuracy of GGA-level for metallic alloys and compounds.