You may explore what your compiler makes available to you by using the selected_real_kind(p, r) function.
The number of atoms in the universe is around 10^80. That, for you, is an infinitesimal quantity! If all the numbers in your calculation are of the order of 10^10^4, simply rescale by that number.
People usually switch to real*16 and such types, if available, in order to achieve extended precision, but you seem to want range, instead.
There are multiple-precision packages available. However, they require significant effort to understand and use. Perhaps, you should explain why you think you need these large numbers, and what you plan to do with those numbers, were they available.
