A form of arithmetic in which numbers are
represented as a fraction and exponent. For
example, 123.45 would be represented as
0.12345 (the fraction) and 3 (because the
fraction must be multiplied by ten to the
power of three), assuming a base-ten system.
In computing, it enables programs to work
with very large and very small numbers, but
is slower than fixed-point arithmetic and
suffers from small rounding errors.