I'm missing something here. Is the point of this system to use the same floating point value in several different places without storing it in several different places? Why not just use a pointer? If the array is fixed-size, where are you storing e.g. the fact that 01110010 is actually 3.79017423523459?