How does forwarding help improve the performance of a pipelined processor? What costs and design issues are involved with doing forwarding? Assume that we have the standard 5 stage pipeline from class and the text. Which, if any, Instructions will actively use all 5 stages? Give an example of one. Give an example of an instruction that will use only 4 of the 5 stages -- which stage does the instruction you have given NOT use