The next year or so is going to be very odd seeing talk about both Intel and AMD. Depending on the angle upon which you view either or both of the companies, just about any opinion, good, bad or indifferent, can be justified.
To keep this simple, let’s take just two perspectives, that of the overclocker, and that of the stock investor.
Overclocker: It’s become pretty clear that Conroe will outperform AM2 chips. What is not clear at all, from the overclockers’ perspective, is by how much when overclocked.
The comparison an overclocker makes is not the comparison CNet makes. Overclockers will want to know by how much an overclocked Conroe will beat an overclocked AM2, and since little more improvement can be expected from 90nm AM2 chips, the question becomes “How fast can you make a Conroe?”
If a low-end Conroe can do 3GHz, this will probably not be something to get all wound up about. People who are way behind the curve or very inclined to upgrade will be more likely to choose Intel, but those with fairly current systems won’t see enough improvement to switch. If it can do 3.5GHz or more, that’s another story, and there will be more switching.
What we really need to see is an Intel roadmap for 2007. If it shows Intel ramping up the speeds bigtime, Conroe is likely to be a killer in the OC ranks. If it doesn’t, it won’t.
The first real shot AMD will have at getting the performance crown back will be the 65nm Revision Gs, and growing indications are they may be some time coming. No doubt we’ll see them first in FXs and the like, and that might occur as early as the end of the year. That 65nm FX ought to give a pretty good indication as to what the cheapies will eventually do, so many will decide what to do based on how it does.
Getting an affordable one from AMD may end up being quite a different matter. If I had to bet now, I’d say that wouldn’t happen until mid-2007, or even later, not too much before Intel will debut 45nm design.
On the whole, this might sound really good for Intel, but keep in mind Conroe and Company represent “Back to the PIII” technology. They pushed the original until it broke, then they broke the PIV, will they break this, too?
Despite its expected performance, Conroe/Merom is still essentially a quick fix. Intel needs a new design, and sooner rather than later.
So, for an overclocker the key timepoints are June and perhaps November/December.
Investor: For someone owning stock, the story of AMD and Intel is one of Ugly and Uglier. The AMD stockholder has to worry about an Intel price war for now. He may or may not be aware of AMD’s delays in getting to 65nm, or just how much juggling AMD will have to do between its two fabs and Chartered, but eventually, he will.
An overclocker would be downright happy if AMD charged less for its CPUs. This would be absolutely horrible to a shareholder, though.
Right now, AMD stock is selling at a price/earnings ratio of 80 (in contrast, Intel’s is about 14). To oversimplify, that means the stock market expects AMD to continue to keep having explosive profit growth like it has the past year, and it’s just not going to happen, even if AMD does pretty well the next few quarters. In the last year, AMD’s profits has benefitted from an increase in ASP due to the wholesale shift from XPs to 64s, a big increase in sales, and a spinoff of an unprofitable memory line. That’s not going to repeat itself in the next year, even if Intel doesn’t do a price war.
To put it simply, if you make 25% more money this year than last, you might think you did well, but if people investing in you expected you to make twice as much, you did badly in their eyes.
(Disclaimer: This is not advice to sell AMD stock, or sell it short, or anything like that. The stock market in the long term eventually gets these things right, but “long term” and “eventually” can mean years. The herd can think one way for a long time despite all evidence to the contrary, then change its mind, suddenly. There are also dark horse factors which could change everything, too, the result of that AMD lawsuit, for instance.)
The AMD shareholder will look downright happy compared to the Intel shareholder, though. The stock has dropped about 30% from the beginning of the year in anticipation of bad news, and it’s going to get it. Its earning release next Wednesday will probably be considered quite bad, and it will get worse in the next few quarters rather than better.
The Intel shareholder may think this is AMD’s fault, and to some extent it is, but the major problem is having less-than-desirable chips to sell, and having to settle for less than usual for them.
Do you think this guy wants to hear about price wars?
An overclocker couldn’t care less if Intel made, say, a billion dollars less profit per quarter. The company isn’t going to go belly-up as a result. For a shareholder, though, that would be absolutely terrible, and it’s likely to stay terrible the rest of the year.
A stockholder couldn’t care less if Conroes show up in June, or how fast they go. All they care about is how fast they restore Intel’s profits, and they won’t have the opportunity to have a major impact on the bottom line until 2007. To a stockholder, owning Intel is like owning a fort being attacked by Native Americans. Conroe in June is like three cavalrymen showing up, with another five a mile away. It really doesn’t help until the rest of the troop shows up, but, on the other hand, the defenders of the fort know help is coming.
In the long run, though, there’s going to be permanent damage left as more people realize that Intel isn’t really a growth company anymore, but rather a company that’s going to struggle to hang on to what it has, and that’s going to be true even if Conroe is good and even if AMD gets beaten back.
Intel is an example of the stock market eventually getting it right. For most of its existence, it’s been considered a high-growth company, and for most of its existence, it was. It hasn’t been for years, though, and it took the stock market a long, long time to realize that.
The CPU industry just isn’t a high growth industry anymore: 10-15% growth a year, and it’s going to be a good deal harder and more expensive to make chips, and it’s going to be a harder sell to get people to replace their computers fairly frequently.
AMD can at least theoretically grow a lot provided they can take it out of Intel’s hide, but when you have about 80% of the market like Intel, how much hide can you grab?
The last number of years, Intel has been trying to squeeze all the profit it can from its revenues (which is probably the real reason why they ran two CPUs designs into the ground), and now they realize they need to expand revenues instead.
Why do you think Intel is so hot on things like VIIV? They’re trying to get people to buy more computers, anyway they can. If Intel (MS, too) thought it could hook up a PC to a blowup doll and persuade you it was . . . a lot better, they’d do that, too. Come to think of it, given the level of Internet porn, that might be a much better idea than VIIV :).
For AMD shareholders, the only thing that will matter in 2006 is whether the profit number keeps jumping up a lot. AMD could have a good, even very good year, and see the stock price drop, and if they don’t . . . .
On the other hand, on the Intel side, it will be initial bad news, and then we’ll see what kind of mood the market is in when the first few Conroe cavalrymen show up.