1. Comparing Performance Characteristics of Two Computing Systems: (5 points)
Computing system #1 has a single issue, scalar, in-order pipeline and a system clock rate
of 3 GHz. It is being used to run a vector multiply-add (axpy) benchmark with a 10,000
element, 32-bit, integer data type vector. The benchmark executes at a sustained rate of
10 cycles per arithmetic operation (CPO = 10). Assume there is no (zero) overhead/penalty
for memory fetches.
Computing system #2 is also a single issue, scalar, in-order pipeline machine. It has a
clock rate of 4.1 GHz, and different architecture which will execute the benchmark at 8
cycles per arithmetic operation (CPO =8).
(a) What will the execution time be for the vector multiply add benchmark when it is run
on computing system #2? Show all the steps used to evaluate the execution time,
and clearly state any additional assumptions you make in your analysis.
(b) What will the speed up of Computing System #2 relative to Computing
System #1 be ? Show all the steps used to evaluate the execution time, and clearly
state any additional assumptions you make in your analysis.