Floating-point

In computing, a floating-point value is a method to store real number in computer memory. The name refers to that the radix point is can vary, unlike fixed-point which has a fixed radix point. All floating-point values can be written as $a\times b^{c}$ for integer values $a$, $b$ and $c$.

IEEE

IEEE specifies multiple standards for floating-point numbers of different sizes. The following table shows the most used:

Name	Size
Half	16-bit
Float	32-bit
Double	64-bit

External Links

• Floating-point on Wikipedia