4. The sign field, the exponent field, and at least the first digit of
5. Each field must be separated from adjacent fields by an
6. All the digit positions in the sign and exponent fields must be
Single Precision Real
This format would appear in memory as:
1-bit sign field
8-bit biased exponent field
23-bit fraction field
A single precision number takes 4 bytes in memory.
Double Precision Real
This format would appear in memory as:
1-bit sign field
11-bit biased exponent field
52-bit fraction field
A double precision number takes 8 bytes in memory.
Note
the mantissa field must be present (any unspecified digits in
the mantissa field are set to zero).
underscore.
present.
The single and double precision formats have an
implied integer bit (always 1).
Floating Point Support
(1 binary digit)
(2 hex digits. Bias = $7F)
(6 hex digits)
(1 binary digit)
(3 hex digits. Bias = $3FF)
(13 hex digits)
4
4-19