I first noticed The Innovator's Dilemma (a superb book I reviewed in November) because it was the cover story of Forbes in February 1999. Andy Grove and author Clayton Christensen were sharing the spotlight, because Christensen was the man who first convinced Intel (Nasdaq: INTC) that AMD's (NYSE: AMD) low-end processors could be a threat. I highly recommend reading that article. It explains why Intel created the Celeron in the first place -- Christensen persuaded Grove to pay attention to AMD.
Unfortunately, the Celeron was too little too late. The revenues from AMD's low-end K5 and K6 lines, minuscule though they were, financed the development of the Athlon. I covered the launch of the Athlon a year ago, and linked to an excellent article from my hometown paper about the then-new chip, but in the end I dismissed it. Sure the chip looked great, but AMD's problem had never been performance, but manufacturing yields. AMD had cried wolf too many times with wonderful designs they couldn't produce in quantity, and by that point I expected another failure and was only listening with half an ear.
I screwed up.
A man who took the Athlon more seriously is Robert X. Cringely, who covered its launch and some of its implications. He pointed out, correctly, that a significant part of Intel's traditional superiority is that it had more breathing room to get things right. Intel had never been in the position of playing catch-up, forced to release a new product before the engineers thought it was ready. Because they were so far ahead, they had time and money to tweak their process and work the bugs out, so they could produce better products cheaper and stay ahead. Their lead was a self-fulfilling prophecy.
When AMD broke out of the box, it leapfrogged Intel and suddenly Intel was playing catch up, and as we saw with Intel's 1 Gigahertz yield problems and their 1.13 GHz processor recall, they don't do much better on the defensive than AMD did. It's not the position you want to be in.
Luckily, Intel had some warning. Andy Grove listened to Clayton Christen, and saw the possibility that AMD would escape from the box Intel had penned it in for so many years. And thus Intel started plans to diversify, way back when.
A month before the Athlon launch, Cringely covered Intel's first big push into networking. He wasn't impressed. I can't say I was either, and a year later we still don't hear much about it. Intel proposed spending a lot of money to become the ultimate e-commerce hosting company. For all I know, they may still be pursuing this. But I can't say I've heard much about it since.
Intel has a ton of money, and there's a ton of opportunity in adjacent market niches. They just have to figure out how to put the right amount of money into the right opportunity. With the Wintel duopoly broken (and trust me, if you work in this industry the cracks have been apparent for some time now), the one true operating system and one true hardware platform are being replaced by commodity interchangeable parts. Increased competition inevitably brings down the margins of any mature commodity. They won't vanish, and Intel's brand name does give them an edge on both volume and selling price, but the downward pressure will only increase in the future. AMD isn't going back into the box, and their success will undoubtedly encourage other challengers (like Transmeta).
The trick is to figure out which way to grow. Flash memory is a profitable niche, currently small but with rapidly growing volume. Intel may still make network hosting pay. And its processor production capacity is still a great asset: if lower margins are inevitable higher volume is still an option in the embedded space or for "Internet devices."
Intel has $13 billion. You can build a new company from scratch with that. Intel's processor division is under siege, but not close to dead. And Intel as a company has plenty of options beyond processors.
If you've been following the "three waves" series I've been doing (also based on Robert Cringely's work), I think Intel just graduated to the third wave.
