NOTE ON FIXED AND FLOATING POINT STORAGE

The bit pattern for a real number numeric type is arranged as follows:

FLOATING POINT STORAGE

Note that assigning more bits for the exponent will increase the range and decrease the precision (and vice versa).

Broadly we can identify two approaches to real number storage:

  1. Fixed Point Storage: Number of bits available for exponent and mantissa always the same (i.e. fixed).
  2. Floating Point Storage: Number of bits available for exponent and mantissa is dynamic (i.e. not fixed).

Floating point storage offers the advantage that it can be used to represent both very large and very small numbers without having to store all the zeroes (although this may involve some rounding up/down). Most modern imperative languages use floating point storage.




Created and maintained by Frans Coenen. Last updated 03 July 2001