here goes nothing. a pipeline is a long string of instructions. these are called stages of instructions. it takes one clock cycle to perform one stage of instruction. with intel, at each of these execution pipelines, they break down instructions into smaller chunks and this results in pipelines that go as long as 30 stages. so it would take 30 clock cycles to execute all 30 stages. with amd their pipelines only has 12 stages (as far as i can remember). that means they only need 12 clock cycles to perform the same amount of instructions. so at each cycle they perform more instruction executions. so in order for intel to keep up, they have to increase the clock rate at which each stage is executed.
heres an analogy i developed. a bakery is called intel and amd. they need some wheat so they call their supplier across the street. intel uses ferraris to ship the wheat and amd uses semi trucks. the ferraris take off at 150 mph across the street and there is 30 of them. while on the amd side they have 10 semis truck that slowly make their way and then all the wheat is dumped at the plant. do you see which one is more efficient? yes the ferraris are going faster but the semis are getting more done at one moment.
hope that helps