Floating-Point Data Types

∟Floating-Point Data Types

This section describes value ranges and storage space sizes of floating-point data types: float and double.

Java supports 2 floating-point data types: byte, float and double.

From the data value point of view, each floating-point data type can represent the following real number values:

float - Supports signed real numbers in range from -10**38 to 10**38 approximately with a 7 significant decimal digits precision roughly.
double - Supports signed real numbers in range from -10**308 to 10**308 approximately with a 15 significant decimal digits precision roughly.

From the data storage point of view, each floating-point data type is stored its values differently:

float - Stores its value in a 32-bit (4-byte) space with: 1 bit for the sign, 8 bits for the exponent component and 23 bits for the fraction component, represented in a binary format described in the IEEE 754 standard.
double - Stores its value in a 64-bit (8-byte) space with: 1 bit for the sign, 11 bits for the exponent component and 52 bits for the fraction component, represented in a binary format described in the IEEE 754 standard.

For more information, read the 'IEEE 754 Standards - "float" and "double"' tutorial from my "C# Tutorials - Herong's Tutorial Examples" book at herongyang.com/C-Sharp/Floating-Point-Data-Type-IEEE-754.html.