Or rather, the evil professor giving me an evil question for the evil final.
Show the contents of registers and special bits, and explain what is happening during the multiplication of the two 32-bit floating point (1 sign, 8 exponent, 23 significant, non-biased) numbers [given on the exam]. Use the exact format as table 10-3 from the text [M. Morris Mano, Computer System Architecture] Explain how the sign is derived. Write the product of the multiplication as a separate answer. Show all work.
Yeah. I had to do two of these by hand, in a tiny blue book, and do a question where I had to explain which of the three methods for handling data dependence [hardware interlocks, oeprand forwarding, or delayed load] was the best - and I had to do it all in two hours with a calculus exam only an hour-and-a-half after I finished.
My prof.
'd the whole class. I knew my spine was tingling for a reason when he said to not be intimidated by the length of the numbers.
EDIT: Just for clarification, without reducing the table down to take account for multiple right shifts occuring in a row to take up less space, multiplication of 2 32 bit numbers creates a table that can become an absurd number of rows long EASILY. Crunching lots of tiny-written binary numbers, confining yourself to the little boxes you made and being able to read your handwriting after writing even 2/3rds of that is a real pain for anybody. >_M
