maximum floating-point precision supported


What is the maximum precision supported for a floating-point number?
I wants to use:

integer, parameter :: R4 = selected_real_kind(P=24)

but I got a message back that it’s out-of-bounds.

Is ther a work around?



We support double precision, i.e. real*8 or real kind 8.
That is about 15 digits of precision.

Our real*16 support is still somewhere off in the distance. You can use 3rd party extended precision libraries.