Processing by numbers

Teenagers are always comparing the processor speeds of their respective PCs. And it’s always the kid with the faster clock that’s got the better box. But can Intel put an end to the practise?

There is safety in numbers, and people and things. And in big wads of money and great big platinum rings – Joan Osborne

Speed isn’t everything. But in the PC game, you’d be hard pressed to find a teenager that agrees with you. My kids and their friends are always comparing the processor speeds of their respective PCs. And naturally, it’s the kid with the faster clock that’s got the better box.

No matter how much you might try to tell them about the merits of a slower processor with a bigger on-chip cache or front-side bus, the message doesn’t seem to get through. If ‘his’ runs at 3GHz and ‘mine’ only runs at 2.5GHz, it’s time for dad or mom to start saving the pennies again.

Thank goodness then for Intel, who recently announced that it plans to renumber its processor line up to include more features than just the clock speed alone!

This renumbering, it says, will allow customers to more ‘intelligently and accurately’ distinguish among individual processors by taking into account a broader set of features that contribute to the ‘overall user experience’. In use, the new numbering scheme will describe the underlying architecture, cache, front side bus, and other Intel technologies. And it’s scheduled to take effect next month for Intel’s mobile processors and in June for its desktop processors.

From then on, processors will be categorised by a 3-digit number such as 7xx, 5xx, or 3xx. This BMW-style number, plus the processor family name, such as Pentium M, will then comprise the overall “processor name’.

A higher number within a processor family can indicate more processor features, more of a specific processor feature, or a change in architecture. For example, there may be a case where the processor number is higher due to a boost in the speed of its front side bus (e.g. from 400 MHz to 533 MHz), or an increase in its cache size (e.g. from 512KB to 1MB), even though the clock speed may stay constant or decrease.

Intel claims that users will be able to use the processor numbers to differentiate between the relative overall features within a certain processor family (e.g. within the Intel Pentium 4 processor family) and within a numbering sequence (e.g. 550 versus 540).

But many are very sceptical. The good folks at the Gartner Group, for example, believe the new numbering will not help Intel achieve its aim and will be largely ignored by PC buyers, sellers and users! Intel’s aim is reasonable, they say, but its execution is flawed. Higher numbers indicate more features, not more performance, and users will inevitably be confused. The numbers don’t explicitly reflect clock speed, and users will have to use look-up tables to interpret them!

I’m a bit more of an optimist. For one thing, Intel’s new numbering scheme can’t help but divert the attention of teenagers away from the rather antiquated benchmark of processor clock speed. And that can’t be a bad thing. Especially if they start discussing the really important critical issues in computer architecture today such as resource utilisation and hyper-threading technology instead.

On the web