The Summer Offensive . . .

Add Your Comments

HKEPC has some summer Intel pricing lists.

Lots of new models will come in, lots will go out, but what they all have in common is that they’ll be cheap by Intel standards.

The most expensive non-Extreme dual-core chip (a 3Ghz) will be just $266, and everything else will $183 or less. Low-end models will approach the $100 level (and probably be less than that on the Dell level).

It’s obvious that Intel will save the high price tags (and likely initial 45nm production) for the quad-cores, and continue to strangle AMD at the lower-end, replacing the Pentiums with C2Ds.

What is Intel out to do? I think Joe Osha from Merrill Lynch has this exactly right:

“The central issue for Intel is always gross margin… We think that Intel is a lot more interested in reducing AMD to its former status than in showing much gross margin improvement this year. . . .” Osha says Intel has “bigger plans” for 2008 on the gross margin line, but warns that “Intel’s plans aren’t going to work out unless AMD backs off.”
Intel’s focus right now, he says, “is putting AMD back in its box, and those efforts are going to take precedence over generating any additional improvements to Intel’s gross margins this year.” Any extra margin, he says, will be passed along to customers in the form of lower prices. Intel’s plans, he says, seem to involve AMD cutting its capital spending, perhaps due to reduced access to the capital markets.

Let’s review what AMD wants to do the next few years. First, it has to finish revving up Fab36 to full throtlle the next few months. Then it plans to gradually take Fab30 to renovate that, using Chartered to make enough (higher-cost) chips to cover any production shortfalls during the renovation, then to increase overall capacity.

Finally, if all goes well, AMD will build another fab in New York State, and roughly around the time that’s finished, that antitrust suit will go to trial and if that goes well, Intel will effectively be hamstrung from doing these sort of things to AMD in the future. (Damages, if any, wouldn’t come for years after that, with appeals and all, but it’s the hamstringing, not the cash, that AMD wants).

At that point AMD ought to have at least two fabs (probably more like 2.5 or more) capable of making current chips at any one time. At that point, we would seriously have to consider AMD to be Pepsi to Intel’s Coke.

Intel is out to stop this at practically any cost. Why? Well, financially, a bigger AMD will inevitably mean permanently lower profits at Intel: fewer sales, lower prices on those sales.

That’s a good enough reason, and it’s true enough, and no doubt will be the reason given publicly by Intel.

But I don’t think it’s the real reason. I think the real reason is all about “putting AMD back in its box.”

I think the real attitude is more like, “We’re the Man in this industry; we’re proud of it, and we’re not sharing.”

Yes, it also conveniently translates to “We’re the (as close as the government lets us be) monopoly, we’re proud of it, and we’re not sharing,” but I guess that’s one of the fringe benefits of being the Man. My point is that Intel’s intent stems from sources much more visceral than beancounting.

And this is what beancounters like Osha don’t get. You see, Osha doesn’t like what Intel’s doing at all. He’d rather Intel call off the price war and be satisfied with higher prices on fewer chips.

No way. This is hormonal, not financial, for both sides. Money is just the weapon. Intel is saying, “I am the Alpha Male. So long as I am strong and you are weak, I will kick sand in your face and beat you down and keep you from getting strong.”

AMD’s attitude is “You have kicked sand in my face and beat me down all my life. This time, I will do whatever it takes to get strong so you’ll never be able to do it again can’t beat me anymore, or die trying.”

This doesn’t mean AMD, the company, has become a suicide fabber. This breakout attempt will stop if/when the money stops, and AMD will be forced to drop its grand plans for something more modest, but if/when it does, so will the careers of the head AMDers. A new team will come in, and things will get quiet for a few years.

So long as Intel has the bigger stick, it’s going to beat AMD into submission whenever it gets uppity. Until AMD somehow grows a bigger stick, or successfully sics the government on Intel, or sells itself to somebody with a bigger stick, it’s going to get beaten into submission, eventually whenever it gets uppity.

It all comes down to testosterone fueled by money.

Ed


Leave a Reply

Your email address will not be published. Required fields are marked *