Unformatted text preview: g any of round towards zero, round to nearest, or round
down as the rounding mode, the correctly rounded result is 215. If round up
is used, the correctly rounded result is (1:00000000000000000000001)2 215.
Exercise 14 In IEEE single precision, using round to nearest, what are the
correctly rounded values for: 64 + 220, 64 + 2;20, 32 + 2;20, 16 + 2;20 ,
8 + 2;20. Give the binary representations, not the decimal equivalent. What
are the results if the rounding mode is changed to round up?
18 Exercise 15 Recalling how many decimal digits correspond to the 23 bit fraction in an IEEE single precision number, which of the following numbers
do you think round exactly to the number 1, using round to nearest: 1+10;5,
1 + 10;10, 1 + 10;15? Although the operations of addition and subtraction are conceptually
much simpler than multiplication and division and are much easier to carry
out by hand, the implementation of correctly rounded oating point addition
and subtraction is not trivial, even for the case when the result is a oating
point number and therefore does not require rounding. For example, consider
co...
View
Full Document
 Spring '09
 LIE
 oating point

Click to edit the document details