Bruce, much of Intel's growth can be attributed to a general boom in the semiconductor industry. 23% growth might not be bad, but AMD's sales nearly doubled. Granted, this was off of a much, much smaller revenue base, but that growth number still speaks volumes. What's interesting about AMD's recent resurgence is the fact that unlike other upstarts that have challenged Intel's position in the past, AMD isn't trying to come after them by means of offering dirt-cheap low-end products at the expense of razor-thin margins. This approach never worked, as the margins such companies attained were always too low, and they could never compete with Intel's massive economies of scale.
But with the Athlon, AMD, on the other hand, is taking on Intel in the high-end, and keeping their prices fairly high. And they're winning nonetheless. While by no means a discontinuous innovation, this is definitely a dramatic turn of events. The Athlon hasn't been released for laptops yet, but it's set to be soon, and AMD could get a leg up on Intel here as well. Here it doesn't matter if AMD is succeptible to the whims of Intel's architecture designs. What matters is whether they can still produce a competitive product anyway, and it seems that they can. And it also appears that AMD might hold onto the lead in the upcoming Thunderbird vs. Williamette battles. Of course, this may just as easily not be the case.
Intel, at least for the time being, appears to have lost the technological lead that they've had in the microprocessor market for longer than most of us can remember. What they have left is enormous sales and marketing clout, and one of the most valuable brands in the world. Still not a bad position to be in by any stretch of the imagination. I remember reading a post at TMF in which an IT guy talked about how he uses AMD's chips in his own computer, but makes sure to buy nothing but Intel-based PCs for his company. That's the power of inertia, that's the power of being the market leader.
But the same things could've been said for IBM's mainframe business in the late-'80s as Unix and Netware servers started to gain popularity. Granted, Intel's management is much more competent than IBM's was back then, but it doesn't change the dynamics of the situation. Intel, like IBM has to push a technologically inferior product in terms of price/performance, and has to hope to get away with it by means of, well, being who they are. History's shown that such an approach can work very well for a long time, but market share starts slipping nonetheless, until the non-technological advantages are wiped away, at which point the losses become a full-blown torrent.
Also worth noting is the coming slowdown in growth for the PC/laptop microprocessor market. Although growth slowdown's stabilized for the time being, I can't see what's to keep it from continuing to accelerate in the future. It's not just that there's a limit to how much processing power people will want, but that there's a significant percentage of the world's population that will most likely never use a computer in its current form. The market for those who will use such computers is getting very close to being fully saturated, and with even business customers shifting to lower-priced PCs, this doesn't bode well for the microprocessor market at large once this recent boom cycle comes to an end.
Now to recap a few of the other markets:
1. Servers - Wintel platforms gaining market share overall, but Unix still dominates the high-end, and should continue to do so for the forseeable future. Sun reported 42% revenue growth, and not all of it was the mere result of industry growth.
2. Flash memory - I can't criticize much here. Intel seems quite solid as the undisputed leader, although AMD's deal with Nokia does show that the lead may not be what it once was. Still, Intel should see phenomenal growth from this market.
3. Network processors - Don't expect too many high-end switches and routers to use Intel's chips. Nearly all of them still rely on proprietary ASICs. Meanwhile, Level One is losing out badly to Broadcom in the ethernet chip market. I'm sure the 50% number included the Giga acquisition. Giga, meanwhile, has its work cut out for it thanks to Applied Micro and PMC-Sierra. Intel's basically growing with the industry here.
4. DSPs - Intel's an also-ran here. TI and Lucent Microelectronics own this market, with ADI and LSI acting as solid competitors. There's a reason that Intel got DSPC at such a low price. Much of this market will be driven by handset sales in the future. With 3G, Qualcomm, which uses proprietary DSPs in its own ASICs, should be a major competitor as well.
Don't take any of this the wrong way. I don't expect Intel to crash and burn. It should see positive revenue growth for a long time, and if they ever decide to do an IPO for their flash memory business, I'll be the first to buy in. It's just that those growth numbers may not be as large as some are expecting. But I've been wrong before, so you never know.
Preparing to get flamed,
Eric |
