faster way to implement 2 power float number without using divider

Question

I have found codes to implement 2 power float number(2^x where x is float), but, all of them are using dividers(/).

below one is one example where they are using divider for calculating it. http://www.machinedlearnings.com/2011/06/fast-approximate-logarithm-exponential.html

I want to have a code where there are no dividers with as much high accuracy as possible.