Execution Speed
In common use, the execution speed is taken to be the clock speed – that is, the speed at which the fetch-decode-execute cycle is completed.
As will become clearer shortly, this isn’t always the best measure of performance, and there are other ways to identify this, but for the purpose of convention, the execution speed is a direct relation to the clock speed.
Read more about alternative performance metrics here.
Clock speed, cache and core count
Additionally, (also in the above link), as the frequency increases, the power draw and heat generated by the CPU also increase.
Again, as already discussed, the cache and core count will also affect the execution speed.
Methods of increasing execution speeds
Following on from above, an increase in any of the three areas will lead to the potential for increased execution speed: although remember the caveat about parallelisation and multiple cores.
Implications of increasing execution speed
Again, as previously mentioned, increasing the execution speed can result in excess heat being generated. This is why overclocking CPUs often requires an investment in heavy duty cooling hardware.
A less-considered implication is that if the CPU is much faster than the rest of the system, it will not be able to work to its true potential: a fast CPU that spends much of its time waiting for data to be sent for processing is no faster than a slower CPU that doesn’t have to wait.