Comment by userbinator

The Z80 is pipelined and thus has a higher latency but also higher throughput. Besides, memory was the bottleneck, in particular instruction fetches, so multicycle instructions made more sense. Related article: https://news.ycombinator.com/item?id=6341137