Cray CRAY-1 - Double Precision Numbers; Addition Algorithm

0 16
10
63
Operand
1
[
0 0
I
4
Operand
2
I
0
0
5
Result
o --- 0
241
\
Floating point reciprocal approxtmation unit - For the floating point
reciproca
1 approximation unit, an i'ncoming operand wi th an exponent 1 ess
than or equal to
200018 or greater than or equal to 60000
8
causes a
floating point range error. The error flag is set and an exponent of
60000
8
is sent to the result register along with the computed coefficient.
Double precision numbers
The CRAY-l does not provide special hardware for performing double or
multiple precision operations.
Double precision computations with 95-bit
accuracy are
available through software routines provided by Cray Research.
Addition algorithm
Floating point addition or subtraction
is performed in a 49-bit register.
Trial subtraction of the exponents occurs to select the operand to be
shifted down for aligning the operands. The larger exponent operand
carries the sign and the shift is always to the right. Bits shifted
out of the register are lost; no round-up takes place.
0.1~ _________ 4_8 ________ ~
discarded
~
•
Figure 3-4. 49-bit floating point addition
2240004
3-24
E
-/'"