HUGE NUMBERS!

[Jaymie]

Hi, I know it's a bit extreme and has probably been asked many times before, but I'm interested in managing large numbers.

I wish to artificially create large ints, much like 64-bit ints on a 32-bit processor. As I understood it, for the processor to manage such a number it would have to split it into two part and process them separately. I wish to do the same but into many parts, and process them separately. -btw, is that how the processor manages it or does it do it some other better way, or have I totally missunderstood it (likely)? :)

I would go about it by creating a 'container' for the number, an array on ints. Then, add the numbers and carry the left overs and add it to the next int in the array. Is this the best way to do it? I am looking for the most efficient way, it needs to be fast; like the 64-bit int.

How is it done exactly with the 64-bit int and alike? And what would the most efficient way of doing it be?

Thanks.

[Galik]

Are you interested in using a third party library for this?

If so this one might fit your needs: http://gmplib.org/

[jsmith]

I would strongly advise using a 3rd party bignum library -- their algorithms will probably be faster than any
you'll write (no offense).