F i n i s a r |
1
2Table 3.16c: IEEE-754 Single-Precision Floating Point Number Format
FUNCTION | SIGN | EXPONENT |
|
| MANTISSA | ||
BIT | 31 | 30……………………23 | 22……………………………………………………………0 | ||||
BYTE |
| 3 |
| 2 | 1 |
| 0 |
ß Most Significant |
|
|
|
|
| Least Significantà |
3
4
5
6
7
8
9
10
11
Rx_PWR(4), as an example, is stored as in Table 3.16d.
Table 3.16d: Example of Floating Point Representation
BYTE | CONTENTS | SIGNIFICANCE |
ADDRESS |
|
|
56 | SEEEEEEE | Most |
57 | EMMMMMMM | 2nd Most |
58 | MMMMMMMM | 2nd Least |
59 | MMMMMMMM | Least |
12
13
14where S = sign bit; E = exponent bit; M = mantissa bit.
15
16Special cases of the various bit values are reserved to represent indeterminate values
17such as positive and negative infinity; zero; and “NaN”or not a number. NaN indicates
18an invalid result. As of this writing, explanations of the IEEE single precision floating
19point format were posted on the worldwide web at
20
21http://www.psc.edu/general/software/packages/ieee/ieee.html
22
23and
24
25http://research.microsoft.com/~hollasch/cgindex/coding/ieeefloat.html.
26
27 The actual IEEE standard is available at www.IEEE.org.
9/26/02 Revision D | Page 23 |