A form of arithmetic in which the decimal point is always in its correct position in relation to the digits. In computing, it is faster than floating-point arithmetic but less suitable for very large or very small numbers.