That's what Math::BigFloat is for. See, as explained above, many languages only support a certain level of precision natively for storing floating numbers. If you need precision for floating-point math, the module above is your friend.