The Myth of GHz part 2
Well I have given a few days. sorry. Medical issues in the family. Nothing as bad as in the past but non the less needed time to take care of it.
But in the first installment I was giving an example of 2 engines and the fact that they are Identical and left it at that.
The important part I left out was that the gear boxes are diffent on the motors and the engine running at 8500 RPM will actually push the car along faster.
So why did this start up. During the Microsoft/AMD Some people were asking when AMD is going to catch up on the GHz range that intel is at.
But the are comparing Apples to Oranges. When Intel Desinged the P4 they designed it primarily to win the GHz race that AMD beat them at. (hey I call them as I see it)
The P4 has some very nice designs to it. It can move Lots of Memory quickly, as a matter of fact it generally needs faster speced memory so that it doesn't become starved. There are several other nicities I won't get into (Intel processor are generally better at encodeing decodeing video either with or without Hyperthreading)
But in that Design change the Gear box was changed. In CPU design terms. the Processor actually perfoms less Instructions per Clock cycle. Some to the tune that P4's when first produced and with the PIII still around under performed them.
In chip design there are always tradeoffs to be made. for heat contraints etc.
While I won't say frequncy doesn't matter, (it does) there are a ton of other issues that must be taken into consideration when considering a Chip. a 3.5 or 4.0 GHz chip may not be better than a 2.5 or 3.0 GHz chip. or it may even be the same when all the aspects of computing are taken into consideration.
As we quickly approach the limitations of Silicon and lithography issues become even more contrained. (each die shrink puts more issues at hand as we get into the quantum instabilities of atoms and electrons)
The next few gens of processors are going to be interesting to say the least.