Apple is one of the manufacturers that is on the forefront of the clockspeed-does-not-equal-performance battle and that's not surprising. Their G3 and G4 CPUs have been stuck at a maximum clockspeed of 500MHz for years it seems. Apple was still promoting their 500MHz machines when Intel and AMD were fighting a fierce battle over who's the clockspeed champion and clockspeed climbed to over 1.5GHz in just under a year's time.
Unfortunately Apple wasn't able to join the race and offer machines with a similar clockspeed thus they resorted to the exact same marketing tools as I mentioned in the introduction to show their customers that a 500MHz G4 compared very well to a +1GHz Intel Pentium III or AMD Athlon. The G4 was blessed with Apple's Velocity Engine, basically a set of SIMD optimizations, that, like SSE and 3Dnow!, could handle multimedia and other tasks faster than a CPU without these instructions. Apple claimed 'Supercomputer' performance for their G4 workstation but all they did was break the Gigaflop barrier with a specific application under very specific conditions, upon closer inspection nothing to write home about.
Never the less the media jumped right on it and Apple sold a lot of G4s just because people thought they bought 'Supercomputer' performance at a bargain price. Marketing and the art of presenting a product properly has always been one of Apple's, or should I say Steve Jobs' strengths. Today the G4 still isn't available with clockspeeds over 866MHz whereas Intel is pushing towards 2.5GHz already. And although marketing can paint a whole different picture of which CPU is actually faster I doubt that Apple will someday claim that their 866MHz G4 is a match for Intel's 2.5GHz Pentium 4.
That they gave it a try though can be seen at the following URL, http://www.apple.com/g4/myth/
be sure to watch the presentation as it is a good attempt to fool the customer into believing their 866MHz G4 is faster than a 1.7GHz Pentium 4 simply by looking at the depth of the pipeline. Their presentation starts off with an explanation about how differences in CPU performance can be explained by the clockspeed (frequency), the number of pipeline stages, the number of functional units (execution unit, fetch, decode, etc.) and cache design of a particular CPU. So far so good, and you'd expect them to touch upon each of the categories separately in order to properly explain the differences.
Unfortunately a few minutes into the presentation they start comparing the G4 to Intel's Pentium 4 and Itanium and the Sun UltraSparc and suggest that their G4 is faster only because it has the shortest pipeline. Now that is a very broad generalization, pipeline depth naturally will influence performance as we explained but the other points mentioned are equally if not more important to a CPU's performance. So what is Apple trying to do? They're simply comparing a 1.7GHz Pentium 4 to an 866MHz G4 and try to demystify the megahertz myth by only talking about a CPU's pipeline, not a word is said about the other architectural differences.
That is a rather shortsighted and clumsy way to tackle the issue, if a short pipeline is all the G4 has to offer over the competition, or is the reason that their CPUs don't reach GHz clockspeeds yet Apple has got some work to do, it is not to be used as an excuse. A short pipeline is often a bottleneck when CPUs try to reach higher clockspeeds and that's something you cannot simply fix with a die-shrink or without re-designing part or whole of the CPU to run at these higher clockspeeds.