A dynamic software float implementation, that allows the user to specify the number of bits they want to use for the significand and exponent. This gives freedom for the user to decide how much precision they would like, versus the space that the float will occupy.