A year ago: Intel shocks, leaks'r'us

Has Great Stan slipped out the details?

Two years ago, we found ourselves in considerable difficulty with Intel because we had managed to see a roadmap which actually took us up to Q2 of 1998. A search on the site will reveal those details. But now it is apparent that Intel, faced with plummeting PC prices and something of a failure to push the PII as much as it would have liked, is actively leaking future information to selective sources. One of those appears to be CNet, which Intel actually has shares in. Anyway, the information about its future plans on prices (at least up until July this year) is out and about, while it has also apparently leaked its roadmap on processors up until the first half of 1999. And as that information now seems to be available to those in the know, we thought we'd share it with our readers as well. (We will have August prices for processors between now and the next issue.)

Intel CPU prices/units of 1,000 (Source: Dr Tom's page)
March April June July
PII/333 $570 $475 $400 $400
PII/300 $530 $370 $300 $200
PII/266 $375 $195 $195 $155
PII/233 $268 $195 $195 not fixed yet
Covington 266 -- $155 $155 not fixed yet
PII/400-100MHz -- $810 $710 $580
PII/350-100MHz -- $610 $510 $415

For the paranoid amongst us, remember that Intel always has several sets of roadmaps and it is adept at throwing dust in the eyes of the world. Slides from these roadmaps are now widely available on the Web, see for example Dr Tom's stuff at

Home Roadmap

and

Business Roadmap

. Of course Andy Grove, CEO of Intel, went on the record two weeks ago about some of Intel's future developments (see separate story), but the roadmap we have seen requires careful interpretation. Look at the maps and prices from the point of view of consumers, for example. If you were contemplating buying a PC now, when it is clear that the PII/233 CPU will drop to as little as $165 in August, would you really want to pay that extra premium now? It is clear from Intel's roadmaps that our predictions as early ago as summer last year are true. The ceramic Pentium MMX is a dead duck. It is not entirely clear whether many of them are still in production - see, for example, the comments Ken Catto made in the last issue about "spot difficulties" with the previous ceramic processors and also the Pentium Pro. Of course, the whole point about these roadmaps and pricing plans is that they seriously impact not only consumers but the entire industry. That, to be fair to Intel, is why they try as hard as they can to keep a lid on them. But even if Intel is not yet listening to its customers, we at

The Register

are. End users are totally unconvinced that a clocking up of a PII, for example, is worth spending the extra bucks on. Intel, itself, is unable to give a convincing tale about why people should upgrade, like they always dutifully did in the past. Maybe it's about time it started to do so...

The Register

also hears that people who have sampled the fabled 100MHz bus from Intel in the shape of the BX chipset, find no real performance gain either. Why should this be so? Answers on a cartridge, please. ®

Sponsored: 10 ways wire data helps conquer IT complexity