arithmetic is implemented as bitwise operations so i’m still confused. the lower the precision / smaller range of values you choose to represent means fewer bitwise operations are required.