If you really need really big arbitrary precision numbers, use Math::BigInt or Math::BigFloat. Or recent perls can be compiled with support for 64bit ints, if your OS supports it.