For a particular format in the IEEE 754-2008 framework, a normal number is a finite nonzero floating-point
number with magnitude greater than or equal to a minimum value where here, is the radix and is the minimum exponent.
Normal numbers can use the full precision available
in a format.
Finite nonzero floating-point numbers which aren't normal are said to be subnormal.
The given definition of normal number shouldn't be confused with the definition of normal number from number
theory. In that instance, a number is said to be normal if it is irrational
and if any finite pattern of numbers occurs with the expected limiting frequency
in the expansion relative to a given base.