Core 2 Duo and Core 2 Extreme: What is Intel Thinking?

In the age of Google Maps and online internet directions it is easy to forget that less than one hundred years ago navigation was a tricky thing. The first trans-American highway wasn’t started until 1912 and the first Rand McNally Road Atlas didn’t appear until 1924. Before those relatively recent achievements, piloting your car (or horse) from one destination to another involved looking for physical landmarks and asking for a lot directions.

Once you start asking for directions on an unmarked road the distances involved become wildly personal. You might hear measurements like “a ways” as in “go past the burnt down barn for a good ways and turn left on the road where Farmer Ogden puts up his scarecrow every spring.” Another metric used in the pre-road mapped US, popular in Missouri, was the “see.” If you were interested in going from, say, Jefferson City to Lohman and asked someone for the route there was a good chance they’d stick you on a rutted lane and tell you to travel straight for about 40 “sees.” A “see” is, of course, how far you could see. For example, if you were at point A, you would look in the direction you were headed. The farthest thing you can make out is an old oak tree. The distance between your current position and the tree comprised one “see.” Once you got to the oak you’d scan ahead, the furthest thing you saw would mark the second “see.” Thus the process would continue.

There are obvious drawbacks to the “see” system of distance measurement. A “see” in the hilly terrain of East Tennessee might only be 40 yards, while in the topographically challenged states like Nebraska a “see” would run from Omaha to Lincoln. However ineffectual the “see” is as measurement tool it did have the advantage of giving, at least, a rough idea of distance. Measuring computer performance with megahertz is a lot like measuring distances with “sees,” it isn’t very accurate but at least you get an idea of something.

If the computer chip followed the model of roads-to paper-maps-to-Google-directions with every generation of new chips, consumers would know a little more about the chip powering the machine. The trouble is that the industry is headed in the opposite direction, instead of clarifying they are obfuscating. Put differently, megahertz (as flawed as the statistic may be) is looking like a pretty solid way of judging computer chips right now. It is as if people abandoned the “see” not in favor of maps but in favor of falling on the ground and emitting unintelligible guttural noises while still expecting the person asking directions to get the general idea.

The problem is Intel’s chip naming scheme and Apple’s apparent adherence to it. At this moment in time the best laptop chip Intel offers is the Core Duo. Monday, Intel announced the names of the next processor revision.

What are the clever names? Core 2 Duo and Core 2 Extreme. The chip names are less telling than the “See” mentioned earlier. Why is the Core Duo 2 any better than a Core Duo? Is it really a core^4? What about the core solo 2 (presumably one of these will roll out). Is that chip one core or two? The Core 2 Extreme? Just how does that chip earn the extreme moniker, is it extremely low power, extremely high performance? And, finally, how in the world can the Core 2 Extreme hope to out duel the Core Duo 2? Sure, Core 2 may be extreme but it is obviously inferior to the Core Duo 2 because the name omits the “Duo,” a superior chip would obviously be the Core Duo 2 Extreme (with a Hyper-Threading Velocity Engine).

It wasn’t always this way. At one time Apple took care of which chip revisions were important and which ones weren’t. For example, The G series of chips weren’t called G series by the manufacturers. The chips had boring names like the 750 GX. There were less than trivial differences between some models, but instead of Core Duo 2 Extreming users to death by telling folks they were buying a computer with the slightly enhanced 750 GX instead of the earlier 750 FX, Apple cleverly waited to spring a new round of chips on people until the difference was substantial. With the G4 it was the addition of the Velocity Engine. The G5, as we all remember, but Steve might like us to forget, was heavily hyped because it was a 64 bit chip.

The naming convention generated a lot of excitement. Once Apple fully adopted the G family moniker, users could get excited about the next number up. If you had a G3 notebook you could hardly wait until Apple managed to shove a G4 chip into a fresh new lappy. If you used a G4 desktop machine, the prospect of G5 was awe inspiring. It wasn’t so much that any G(n+1) would make an incredible amount of difference in your day to life, rarely did that occur, it was more about the excitement of being on the cutting edge. The fact that Steve Jobs could talk about a Pringles chip and make it sound as though your current machine was archaic if it didn’t ship with at least a dual Pringle didn’t hurt.

The G naming convention might not have actually meant anything intrinsically, but it was a sign to consumers that their machines were seriously outdated. Sure the difference in performance between a 400 MHz G4 and a 1.62 GHz G4 was probably greater than the difference between the late G4 and early G5 chips, but let us not quibble here—the G5 was a full digit better than the G4 and that was enough to get folks salivating.

With the names Intel is throwing around, the days of breathlessly anticipating the next model may be gone. If Apple had stuck with IBM, the Mac sphere would be buzzing with rumors about the G6 or the Cell processor, and the day when Apple might bust through the 3 GHz barrier. Now we are stuck in a weird place, with no major increments spoonfed to us by the Gx-naming-convention. We suddenly find ourselves face to face with reality: steadily faster machines, but no quantum leaps. Sure, the current state is more reflective of computing reality, but it is also much, much more boring.