Why is it that the smaller instruction set of RISC architecture's does not necessarily contribute to a lower CPU time when compared to CISC?

Question

Recently had this T/F question on a Comp. Systems quiz:

Consider the CPU time formula:

CPU Time = IC × CPI × (clock cycle time).

If we only compare the first term IC, RISC performs better.

And the answer was false. Can someone explain why this is? I thought since RISC has fewer instructions than CISC that the IC on RISC would be lower leading to better CPU time.

score 3 · Accepted Answer · answered Dec 08 '19 at 04:14

3

IC is Instruction Count. It does not mean "how many instructions the CPU implements" but "how many instructions it takes to implement a given algorithm".

Since the instructions in a RISC machine tend to be simpler than instructions in a CISC machine, you need to execute more instructions to achieve your desired ends.

i.e., on RISC, IC is higher and therefore worse (but of course we expect a lower CPI and cycle time to make up for it).

answered Dec 08 '19 at 04:14

Not sure if the cycle times on RISC architecture are faster, but indeed all the instructions should run in one processor cycle, so CPI should be low (https://cs.stanford.edu/people/eroberts/courses/soco/projects/risc/risccisc/). – Hemaolle Dec 08 '19 at 04:28

Why is it that the smaller instruction set of RISC architecture's does not necessarily contribute to a lower CPU time when compared to CISC?

1 Answers1