Original URL: https://www.theregister.com/2000/03/07/intels_little_1ghz_mysteries/

Intel's little 1GHz mysteries…

Every dog has its day in the chip world

By Mike Magee

Posted in On-Prem, 7th March 2000 07:03 GMT

Some signs of a Dell page advertising 866MHz and 850MHz Pentium IIIs appeared and disappeared this morning even before we could tuck into our hot, piping, Ready Brek. Every dog has its day, and while yesterday was AMD's, we are now of the strong opinion that tomorrow will be Intel's, as it retaliates against the coverage that its smaller competitor got yesterday. The share price of AMD went, so to speak, through the sound barrier, closing at $47.5 yesterday evening (cough). With little more to substantiate this apart from nods, winks, nudges, the Houston Chronicle, and piecing everything together, we now believe that Intel will cajole Compaq, HP and Dell to get on its own 1GHz bandwagon tomorrow, as it rolls it some of those and also the 866MHz and 850MHz Pentium IIIs that were lost on the 27th of February somewhere in the equivalent of the Bermuda Triangle of Chips. Intel will also mount a counter-offensive against claims that a 1GHz processor is necessarily faster than its own offerings anyway. It may take pre-emptive action on pricing too. It's perfectly capable of doing so, in what's fast turning from a game of straight-faced poker into a game of Snap! It is likely to cite benchmarks, for example these ones from PC World, which show that a Gateway running a 1GHz Athlon only very, very slightly outperforms a Gateway machine running a 800MHz Pentium III. The reason? Coppermine processors have an integrated L2 cache running at full speed, and a wider pipe between the L2 and the core. The K7 runs the L2 at one third the speed of the core, slowing performance. Later today, we will do a hardware roundup and look at all the reviews we can find of the K7 1GHz. These technical issues of course will be lost on the vast majority of people, reading accounts of announcements from chip companies major and minor, laundered through the press release mill and re-gurgitated wholesale. And if Intel spinolas should start to talk about chips faster than the speed of light, quantum leaps, or the Coppermine being the biggest aspidestra in the world, they know what treatment to expect from this organ... Intel should, of course, have thought of all this before the company let the Megahurtz Genie out of the Klein Bottle, telling consumers that raw speed is a good thing. But that was so long ago, few people now remember and it has to carry on playing the game. ® See also AMD suffers Gigglehurts hype attack AMD puts Intel on 1GHz ropes -- confirmed