You have pipelines in CPUs, not motherboards. Basically, it's chopping the CPU up in different "pipeline stages" to allow for better/more efficient utilization of different parts of the CPU.
Say, for example a CPU is a 100-metres long road that someone (an instruction) can travel along - but you can only have one person on this road at a time. Instead of waiting for the person to run all the way to the end of the road the read is divided to, say, 10 chunks (pipeline stages) that can all have one person walking on them at a time, so you can send a new person on his/her way as soon as the previous one has walked only 10 metres. This significantly increases the processor's throughput.
