Quiz submission record for quiz3-2-1 at Wed Jul  7 01:27:48 2004:

Your Answer for Question 1:
someone could exploit this by storing say an instruction into an int and then when memory loads it, it won't know what to do with it.  or if an int was given a float then the precisions would be off.  one possible technique could be to type check by letting some bits distinguish the types.  however this will use up much sought after register space.

Your Answer for Question 2:
subnormal is when a number that is not shifted right enough so that there is only one digit before the decimal point.  we care about this because we want to normalize numbers in order to be able to add, subtract multiply and divide them.

Your Answer for Question 3:
0.5 has more bits so when you do 0.5 - x the bits at the end of the number won't fall off like it would if you did 1.0 - x

Your Answer for Question 4:
using the standard would allow our coprocessor behavior to be understood by almost everyone.  but the ieee 754 standard have features such as directed rounding and humane exception handling that is difficult to make available in compilers.


Your unique submission ID is quiz3-2-1-cs61c-ao-1089188868-1148.