For the first time since the days of the 386, 1998 was the year that Intel finally found itself under pressure from its rivals. That pressure -- mostly concentrated in retail and low-cost PC markets -- is clearly a target Intel aims to address judging by the news that it will release a series of processors early next year. The most interesting introductions are expected to expand the scope of Intel's mobile and entry-level desktop lines: precisely the areas where it is most under siege. Cyrix, AMD and newcomer Rise are only now making concerted stabs at the mobile sector. On the more important desktop front, reports in the US consumer market make nasty reading for Intel executives and suggest that non-Intel brands have, taken together, achieved a strong leadership position.
In the UK too, AMD and Cyrix have between them had big design wins with the likes of Compaq, lBM, Hewlett-Packard, Packard Bell and Time signing up. About a third or more of UK retail is now in the non-Intel slice of the pie.
Credit to AMD and Cyrix who, after years of playing a Rodney Dangerfield role in microprocessors, have finally succeeded in matching a coherent strategy and product line with the ability to manufacture in volume.
That said, the most worrying aspect for Intel must be that the branding machine that made a household name out of a made-up word -- Pentium -- has for once gone awry. Celeron has had bad vibrations ever since the first cache-less parts failed to match up to any sensible price-performance index. Celeron is healthier now but Intel still needs to differentiate between it and higher-margin Pentium II chips. Nobody in the industry will be too surprised, therefore, that it ploughs on with a 66MHz front-side bus.
The failure to release chips based on an `MMX2' set of multimedia instructions to compete with its rivals' 3DNow! has also been an uncharacteristic failure and provided AMD in particular with terrific momentum among gamers and savvy buyers. The Katmai processor will be a very important one for Intel.
And despite naïve suggestions to the contrary, Intel cannot just buy itself out of trouble. Its current price list is low enough as it is, but Intel's duty to its shareholders precludes the prospect of a total bloodbath.
But nobody should cry too much for Intel just yet. Despite the odd profit warning, Intel remains a powerhouse which is several times larger than its rivals and hugely profitable to boot. 1999 should be interesting.