GE-635

Floating Point:

Single Precision (one 36-bit word) floating point :
GE-635 Floating Point Representations: Single Precision (word 1 of 1)
0
 1
 2
 3 4
 5
 6
 7
 8
 9
 10
 11
 12 13 14 15 16 17 18 19 20
 21 22
 23 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
Sign Bit Exponent (no bias) Sign Bit Fraction: bit 9 is the most significant bit

Double Precision (two 36-bit words) floating point:

GE-635 Floating Point Representations: Double Precision (word 1 of 2):
0
 1
 2
 3 4
 5
 6
 7
 8
 9
 10
 11
 12 13 14 15 16 17 18 19 20
 21 22
 23 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
Sign Bit Exponent (no bias) Sign Bit Fraction: bit 9 is the most significant bit


GE-635 Floating Point Representations: Double Precision (word 2 of 2):
35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Fraction (Second, less significant part): bit 35 is the most significant bit in this word

EXPONENT (E): To the base 2, with no bias.  

FRACTION: Has no hidden bit.

NORMALIZATION: Floating point numbers may be either normalized or not.  Note that unnormalized numbers may have a very significant loss of precision.

ZERO: Represented as 0*2-128

INTEGERS: In the GE-635, integers may be held in half words of 18 bits each, whole words of 36 bits, or in doublewords of 72 bits.  They are in 2's complement.  Integers may be interpreted as "Integral" or "Fractional".  Integral integers have the binary point (like the decimal point) to the right of the least significant bit.  This is the same as most other computers.  Fractional integers have the binary point to the left of the most significant bit.

In addition, both integral and fractional integers may be either in the "algebraic" or "logic" group.  Algebraic integers are conventional; the leftmost bit is a sign bit, a "0" indicating a positive number, and "1" indicating a negative number.  Logic numbers are unsigned, positive binary numbers (which may be fractional).