Integrated DOS and the total number of (valence) electrons

[terencelz]

In the manual page for DOSCAR, it says (the integrated DOS) conserves the total number of electrons exactly. I'm not sure if it's talking about the "total" electrons, or just the valence electrons. For example, I have a two Si primitive cell, and DOSCAR gives me 16 as the result of the integrated DOS. But each Si has 4 valence electrons, meaning 8 in summation.

And, from what I learned, to get the total number of valence electrons, one needs to integrate the product of Fermi distribution and DOS. Since the temperature is far lower than the fermi temp, when E>Ef there should be very few electrons. But apparently from the plot of DOS there are a lot of states, and they are taken into account to get the number of 16.

Any help is appreciated!

[terencelz]

I guess I should be looking at the Integrated DOS before the fermi energy because the Fermi distribution is 1 before it. And it does give me 8.