To add: with pipelined processors, it doesn’t make sense to list “clocks per instruction” as a single number. Since the advent of superscalar processors, it mak

To add: with pipelined processors, it doesn’t make sense to list “clocks per instruction” as a single number. Since the advent of superscalar processors, it makes even less sense. Hence why there’s “latency” and “throughput”.