Computers as Components, 2nd Edition

128 CHAPTER 3 CPUs

Inspection of the code shows that the only instruction that may take more than one cycle

is the conditional branch in the loop test. We can count the number of instructions and

associated number of clock cycles in each block as follows:

Block Variable # Instructions # Cycles

Initiation t

init

77

Body t

body

44

Update t

update

22

Test t

test

2 2 best case,

4 worst case

The unconditional branch at the end of the update block always incurs a branch penalty of

two cycles. The BLT instruction in the test block incurs a pipeline delay of two cycles when

the branch is taken. That happens for all but the last iteration, when the instruction has an

execution ...