Reserve Memory for GMP Integers?

I'm working on an algorithm which calculates the square root of 2 to many decimal places. I'm iterating a pair of integers which get very large, and I think I have some decent speed but I want to reserve memory so that GMP doesn't have to keep re-allocating as the integers get larger.

Does anyone know how?
It's always important to RTFM.
I have read the relevant sections of TFM and strangely enough, there's a way to do this for floats but no equivalent way for the other data types.
Topic archived. No new replies allowed.