The largest float could be represented as a 128-bit integer consisting of
24 ones followed by 104 zeroes.
N = (2^24) - 1 in base two is 24 ones. So the largest float is
N * (2^104), i.e. N with another 104 zeroes on the end.
In base ten that's about 3.402823466e38.
The 'most negative float' is the same number times -1.
Here's the relevant #define from FLOAT.H for Visual C++
24 ones followed by 104 zeroes.
N = (2^24) - 1 in base two is 24 ones. So the largest float is
N * (2^104), i.e. N with another 104 zeroes on the end.
In base ten that's about 3.402823466e38.
The 'most negative float' is the same number times -1.
Here's the relevant #define from FLOAT.H for Visual C++
#define FLT_MAX 3.402823466e+38F