What AMD talked about
AMD released its financial information, had some good news, had some bad news.
The good news is that AMD slightly beat the earnings estimates, sold all its 3.6 million TBirds/Durons, and is optimistic they can sell all the TBirds they make next quarter.
The bad news is there aren’t cheap socket A mobos around, and won’t be for the Christmas season. AMD has been forced to drop Duron prices way down to compensate for that. This was part of the reason why AMD only got $90 a CPU rather than $99, as they expected. They have some doubts they can sell all the Durons they can make due to this.
AMD also doesn’t anticipate any big increases in profit for a while.
What AMD didn’t talk about
Durons haven’t been the only thing AMD’s been discounting lately. Take one look at the latest TBird pricing I posted on the front page today (10/13/00).
AMD said they made close to 7 million chips last quarter. They said they’d make about 9 million this quarter. They said more were coming as Dresden continues to ramp. They said they figured about a 20% marketshare for the fourth quarter, with 43-44 million CPUs sold in the quarter.
They didn’t mention that Intel will be up to eight fabs making Coppermines by the end of the year, and what they really didn’t talk about is what happens when Intel can make 30-35 million chips, and they can make ten, but the world only wants 30-35 million, like they will the first few quarters after Christmas.
What it looks like AMD is trying to do
AMD wants to clear out “old inventory” this quarter. After all, they are trying to double CPU sales this quarter, and nice cheap computers for Christmas is a good way to make more people feel warm and fuzzy about them for the future.
K6-2 sales are being phased out, so while selling Durons for $45 is aggravating; it’s no worse than selling K6-2 for that price or below. Selling TBirds for $120-150 is not ideal, either, but it’s still better than the $90 ASP they got last quarter. In any case, AMD is ending production of the low end Durons and TBirds anyway.
Getting On Top of Intel
It will no doubt be intensely aggravating to Intel to be undercut so much, but since the Christmas quarter is going to be a good one, Intel was probably resist cutting prices as much as possible since there will be more than enough demand for their CPUs this quarter.
Next quarter is another story.
By early next quarter, the entry-level Son of Duron chip will probably be 850Mhz, and Son of TBird will be 900Mhz or even 1Ghz. Entry level.
Sure, Intel will have Willamette, but they aren’t going to be making very many.
After that, what does Intel have? What they have right now. If AMD charges $80 for a 900Mhz Son of Duron, and $150 or even less for a 1Ghz processor, that’s an improvement for AMD. What is that going to do to Intel? How much can Intel charge for any PIII?
One of two things will happen: Intel either fights on price or it doesn’t.
If it does, we have a price war, and you may get your next CPU from a cereal box. If it doesn’t (more likely), then AMD’s going to sell all its chips, and Intel won’t. Either way, Intel’s going to make a lot less money for at least the first quarter and probably well into the second. If you think Intel had bad news this quarter, just wait six months.
What the analysts see and don’t see
The analysts see dropping prices, and not quite making projected ASP, and they think the sky is falling. What they don’t realize is that AMD should make at least a bit more money as the product mix goes to more expensive processors. What AMD enthusiasts don’t see is that the shift upward probably won’t be all that much, and could even go downward if Intel decides to duke it out.
However, AMD should be able to keep making about what they’re making now, maybe somewhat more. The analysts don’t like that. They want to see companies making a lot more profit as time goes by, and AMD probably won’t.
However, what they forget is that AMD is now so undervalued that the stock price should go up even if they don’t make much more money the next couple quarters.
If AMD Doesn’t Go Up, Intel Should Go Down
You can make a decent argument for AMD’s stock price staying low because of uncertainties as to how much money they might make if a price war erupts.
What you cannot rationally do is say that AMD’s stock price should stay low and Intel’s should stay high. Intel can’t wallop AMD without walloping its bottom line even more. How many Itaniums do you think Intel can sell? Even if Intel doesn’t wallop AMD, it’s going to lose a lot of profit from reduced volumes.
There is no problem with computer demand. We’re still getting decent growth, and should have a very good Christmas quarter. Any stock analyst who babbles about that is clueless.
The real problem is going to be a glut of processing capability in 2001. That’s the reason why CPU stocks shouldn’t be looking all that good right now in general, and Intel in particular. AMD has the chance to do OK, Intel doesn’t.
And that’s what the analysts are blissfully unaware of. Until sometime next April.
Be the first to comment