Forty years after it started toiling over microprocessors for devices as epoch-defining as (yes) the digital watch and pocket calculator, Intel has reached a new fork in the road of its own skilful creation.

On the face of it, using the Consumer Electronics Show (CES) to big up a few new chips is nothing to get excited about. The Core i5 and i7s families aren’t new, and the unreleased Core i3 has been well trailed in the press, so is there any point in re-straining the tea leaves?

Intel’s ‘sort of’ announcement is that it can now make the chips using a 32nm process, gobbledigook to consumers, but which does reduce heat, allowing higher frequencies and reducing energy consumption. This year will see four of Core i3 chips, plus eight versions of the well-received Core i5, and a further five Core i7s.

Because of all this heat and energy reduction, many of these will end up in laptops and other mobile computers, and not just in the desktops fewer and fewer of us are buying.

Look at the specs more closely, and what is going on is a timely shift away from the previous emphasis on ‘cores’ to hold the marketing story, and for a good reason. Intel will doubtless add more cores to its chips in the coming year or two, but everyone knows that a law of diminishing returns sets in even at 4-core level. Operating systems and applications just can’t make efficient use of such hardware right now, unless you’re using a specialised application such as a video CODEC or software compiler.

Hey Presto, Intel has cleverly reinvented the ‘core’ story with the ‘i’ moniker. The supposedly new Core i3 actually is actually just a fancy two-core chip with and the Hyperthreading (but no TurboBoost) that first appeared on the later Pentium 4s as long ago as 2004. Whether it offers much more than today’s standard Core2 Duo chips is a very open question, but it certainly FEELS new. Laptop makers will lap it up.

And the Core i5, previously an amped 4-core chip? That now also comes in new 2-core versions, so the family has two quad-cores on 45nm fabbing, with four dual-cores on a 32nm process.

Intel has been accused of confusing the heck out of ordinary consumers, once used to understanding chips by their megahertz rating. That was always a bit of a con. Then we got chip names on their own - the Pentium 4 - which made things confusing in a different way.  Then came the ‘core’ story. Now we have...what? Three chip families based on established two and four-core technology that has some startling innovations (see the boost to integrated graphics power) but is also in danger of turning into yet another silicon marketing exercise.

All the consumer is supposed to do is distinguish between the entry-level (Core i3), the performance (Core i5) and power-at-any-cost (Core i7). At a deeper level, these are, in fact, three different platforms, each coming with a different chipset and socket number, each having a varying number of cores activated. In the case of the new Core i5s, that really just means that the chip is a quad-core with two cores turned off.

Confused? I am. Suspect that Intel is pulling another marketing trick to exaggerate the role of brute processing power for the vast majority of consumers? Make your own judgment and then ponder whether the dual core or Core2 Duo chip inside the average PC is really that obsolete after all.