[josefx]

You can use doubles to store and calculate exact integer values. You just wont get 2^64 integers, instead you get the range +/-2^53 .