How can low latency not = high throughput?

Question

The graph above shows the latency of read and update operations while the column shows the throughput(operations/second) I am unsure how MyISAM can perform operations with such low latency but still be level with InnoDB in terms of throughput? MongoDB also dominates in terms of throughput but has a higher latency than MyISAM.

How do these results make sense?

Answer 1

Well, say you have a formula 1 car and a bus. The formula 1 will quickly carry a single pasenger while the slow bus can cary say 40 passengers at a slower pace. The bus will certainly achieve a higher throughput but with a higher latency. Note however, that the bus will lose its advantage in throughput if it is used to carry one passenger at a time.