Intel released their financial data yesterday, and, unrealistic stock market expectations notwithstanding, they did fine.
Intel keeps pumping out a bit over $5 billion worth of CPUs a quarter, and makes about $1.4-$1.5 billion profit from them (other Intel divisions generally lose money). They’ve pretty much been doing that the last two years.
Making $6 billion profit a year on CPUs looks pretty good until you remember that Intel was making about $12 billion a year in 1999-2000.
And even $6 billion a year doesn’t look so hot when you realize Intel has been spending a good deal more than that on new facilities. Whether it’s $6 or $6 billion, you can’t spend more than you can make forever.
Intel looks upon this as an investment. It’s cloudy now, but the sun will come out some day, and then Intel will make hay when it starts shining.
This makes a lot of sense if this is just a slump from business-as-usual. But is it?
Where The Buyers Are and Aren’t
Perhaps the most ominous indicator that this is not just a slump comes from where Intel makes its money in the world.
Apply Intel’s revenues to the percentages and you get:
Intel’s revenues in 2000 were 25% higher than in 2001/2002. Most if not all of that was due to lower pricing. If you sell ten items at $2 in Year One, and sell ten items in Year Two for $1; your revenues drop 50%, but you’re selling the same number of items.
So, to get a more accurate approximation of unit sales (this is like adjusting for inflation, just the other way around), let’s adjust the 2000 revenue figures downward 25%:
Even after you adjust the 2000 figure for the 25% drop in revenue between 2000 and 2001/2002; the American market still shrunk about 25% between 2000 and 2002.
In contrast, the Asian-Pacific figures (once revenue-adjusted), show decently healthy growth.
The Intel figures really aren’t static at all. What’s happening is that East Asian revenue growth has been masking American revenue decline.
It’s not like Asia-Pacific is booming while the Americas are in depression, either.
What’s the difference between the two regions? I think it means the Americas (i.e., the U.S. and Canada) are pretty much saturated with computers while Asia-Pacific isn’t.
(European and Japanese revenue figures show a static-to-declining trend, though decline is not to the same degree of the Americas)
Temporary Or Permanent?
Up to now, the conventional wisdom in at least the tech sector has been that this slump is really just a slump, and once the slump was over, the computing market would go back to 15% growth as usual.
What is more likely (and a growing opinion outside of the tech-glazed world) is that for the medium to long-term, the gravy train is over. The American computer market will stop declining, and eventually perk up some, but the days of perpetual 15% growth are over. For good. To a lesser extent, that will be true for Europe and Japan. The computer markets in those regions will settle largely into “replace the old ones,” with long-term growth rates more like 5%.
East Asia looks to be pretty much the only significant (in volume terms) growth area for some time to come. Even there, you have to wonder how long or well the “Let’s give Intel $150-200 for a CPU” tune is going to play.
If it’s not even playing horribly well in the U.S. anymore as the product mix goes to cheaper and cheaper processors, it’s hardly going to do better among poorer countries and poorer people where cost is inevitably a huge factor.
Whatever is said about Intel goes double for Microsoft
If you look down the road five to ten years, Intel’s real competition will probably not be AMD (which from this perspective has the same bigger, faster, better engineering mindset as Intel), but rather Via (or other companies following their path).
If nothing else, AMD will be remembered for cracking Intel’s pricing structure, but at this stage, I don’t think it even matters much anymore from a pricing perspective long-term whether AMD even survives or not. If AMD died tomorrow, Intel would certainly try to raise prices, but I don’t think they can go back to the good old days anymore. The demand for low and lower prices has reached the point that if Intel tried to turn back the clock, Via would suddenly get very popular, just like they did when Intel tried to make this an RDRAM world.
If AMD does survive, what the computer historians may well record will be two dinosaurs mutually assuring each other’s destruction by competing using the old parameters and completely ignoring the Asian “good enough and really cheap” competition sneaking up from behind.
Want to see a glimpse of what’s sneaking up from behind? Take a look at the Walmart $200 Via/Linux PC.
Intel seems to be planning on more of the same indefinitely: more expensive fabs making faster and faster processors for about the same (considerable) amount of money.
In the long-term, that may be a losing strategy.