numbers - technical definition

In a computer, numbers can be stored in several forms. Although they are all coded as binary digits (bits), BCD and packed decimal numbers retain the decimal relationship of a number, whereas fixed and floating point do not.

Binary Coded Decimal (BCD)

BCD encodes each decimal digit in a single byte. The number 7100 would take four bytes. A variation, called "packed decimal," encodes two digits in one byte.

Binary Fixed Point

This method converts the entire decimal number into a binary number, placing it in a fixed unit of storage. The number 7100 would require at least two bytes. Binary numbers are calculated faster than decimal (BCD) numbers.

        Bytes  Bits   Values
        1      8      0 to 255
        2      16     0 to 65,535
        4      32     0 to 4,294,967,295

Binary Floating Point

Floating point allows very small fractions and very large numbers to be maintained and calculated quickly. Both the mantissa (significant digits) and the exponent (power to which the base is raised) are converted into binary numbers. See floating point.