Cray CRAY-1 - Page 61

Floating point add unit - A floating point add range error condition is
generated for scalar operands when the larger incoming exponent is greater
than or equal to 60000
8
.
The floating point error flag is set and an
exponent of 60000
8
is sent to the result register along with the computed
coefficient, as in the following example:
60000.4 Range error
+ 57777.4
60000.6 Result register.
Floating point multiply unit - In the floating point multiply unit, if
the exponent of either operand is greater than or equal to 60000
8
or if
the sum of the two exponents is greater than or equal to 60000
8
,
the
floating point error flag is set and an exponent of 60000
8
is sent to
the
result register along with the computed coefficient.
An underflow condition is detected when the sum of the exponents is less
than or equal to 177778 and causes an all zero exponent and coefficient
to be returned to the result register. However, if the sum of the
exponents
is 20000
8
and a normalizing left shift occurs, an exponent of
177778 is sent to the result register along with the computed coefficient.
Underflow is also generated when either, but not both, of the incoming
exponents
is zero. Both exponents equal to zero is treated as an integer
multiply and the result is treated normally with no normalization shift
of the result allowed. The result is a 48-bit quantity starting with bit
16. When using this feature, consider the operands as 24-bit integers
in
bits 16 through 39 even though they are actually fractions with the
binary point
between bits 15 and 16. In the following example, operand
1
is 4 ana operand 2 is 5 to produce a 48-bit result of 24.
2240004
3-23
E