the rock-salt structure
The rock-salt structure is the pattern of ordinary table salt, and it is the simplest and most common structure for an AX ceramic, one cation for one anion. Imagine two identical face-centred-cubic grids, one built from anions and one from cations, slid into each other so that each cation sits exactly halfway between anions along every cube edge. The result is a perfect three-dimensional checkerboard where every ion is surrounded by six of the opposite kind.
Read it as close packing: the anions form a face-centred-cubic (cubic close-packed) array and the cations fill every one of its octahedral holes. Coordination is therefore 6:6, six anions around each cation and six cations around each anion. There are four formula units in the cubic cell (four cations and four anions). This arrangement is favoured when the radius ratio sits in the octahedral window, roughly 0.414 to 0.732; for MgO the ratio is about 0.51, comfortably in range.
A huge family of ceramics takes this structure: MgO (periclase, a workhorse refractory), NiO, CoO, MnO, FeO, CaO, and even many carbides and nitrides like TiC and TiN, where metal and non-metal alternate. Honest caveat: several rock-salt oxides are markedly nonstoichiometric, the classic case being wustite, written FeO but really Fe(1-x)O with missing iron and some Fe3+ to keep charge balance, a reminder that the tidy 1:1 formula is an idealisation.
Periclase, MgO, is rock salt through and through: an FCC oxygen array with a magnesium in every octahedral hole. Its high melting point (about 2850 degrees C) and simple, strongly bonded structure make it a mainstay refractory for steelmaking furnaces.
MgO: FCC oxygen, cations in all octahedral holes, 6:6.
The rock-salt structure is not body-centred cubic. It is two interpenetrating FCC lattices of unlike ions; calling it BCC (which requires identical atoms at corner and centre) is a common beginner error.