Univac 418

Floating point numbers for Univac 418 computers are stored in either two or three 18-bit words.

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
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
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
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
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
Low-order fraction

The exponent is biased by 16,38410.  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.