If you're doing division and multiplications with fixed point arithmetic and the result must fit in a same size word as the operands, you need to decide for eac

If you're doing division and multiplications with fixed point arithmetic and the result must fit in a same size word as the operands, you need to decide for each operation how many fractional bits from each operand you're going to retain in the result. In other words: you need to understand the range of numbers that you're dealing with every step of the way. That's not really an improvement... (I wrote about this here: https://tomverbeure.github.io/rtl/2018/11/26/Racing-the-Beam-Ray-Tracer.html#from-floating-point-to-fixed-point https://tomverbeure.github.io/rtl/2018/11/26/Racing-the-Beam...)