The following is the representation of a 2-word floating point number:
Univac 418 Floating Point Representations: (Word 1 of 2) | |||||||||||||||||
17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
Sign Bit | Exponent | High-order fraction |
Univac 418 Floating Point Representations: (Word 2 of 2) | |||||||||||||||||
17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
Low-order fraction |
The exponent is biased by 128. Negative floating point numbers are the one's complement of the positive number.
"Double precision" floating point numbers are as follows:
Univac 418 Floating Point Representations: (Word 1 of 3) | |||||||||||||||||
17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
Ignored | Exponent |
Univac 418 Floating Point Representations: (Word 2 of 3) | |||||||||||||||||
17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
Sign Bit | High-order fraction |
Univac 418 Floating Point Representations: (Word 3 of 2) | |||||||||||||||||
17 | 16 | 15 | 14 | 13 | 12 | 11 | 10 | 9 | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
Low-order fraction |
The exponent is biased by 16,384_{10}. Negative double precision numbers are complements of the positive numbers; however, only words two and three are complemented. Bit 17 of word 2 is the sign of the fraction.
Integer numbers are stored in one 18-bit word. They
are in sign-magnitude format.