There are two floating point formats: short and long
. Short is shown below:
Xerox/Scientific Data Systems Short Floating Point Format  
S  Exponent  Fraction  
































Long floating point is the same, with the addition of another word (32 bits) of fraction, but see RANGE, below. The first 32 bits are the same as above.
EXPONENT (E): To the base 16, with a bias of 64.
ZERO: A positive floatingpoint number with a fraction of zero and a characteristic
of zero is a "true" zero.
A positve floatingpoint number with a fraction
of zero and a nonzero characteristic is an "abnormal" zero.
NEGATIVE FLOATING POINT NUMBERS: Same as positives, but in two's complement.
RANGE: The value of the number N is N = F x 16^{E64} , where
Note that the lowest nonzero values of F may have significantly reduced precision!