It is noted that the technology for the manufacture of ferrosilicon-magnesium modifiers (crystallization into an ingot in a mold or into a chip on a water-cooled drum) forms their quantitative and qualitative structural differences: the size and distribution of phases, the scatter of the content of elements, etc. The fundamental differences are shown and the features of the distribution of silicon as a base element in ingot and chip modifiers are described. It was found that in ingot modifiers the vast majority of silicon-containing phases are ξ(FeSi) and ζ' (FeSi2), and with a significantly lower content of pure silicon particles and phases, these are α1(Fe3Si) and α2(Fe2Si). The picture of the phase structure of the modifier chip has been determined. In it, the 1st mode corresponds to the presence of a set of phases α1(Fe3Si) and α2(Fe2Si), and the 2nd mode, as well as in the ingot modifiers ξ (FeSi) and ζ'(FeSi2), but is shifted to the lower boundary of the region their existence. Particles of pure silicon are completely absent in the chip modifier.