AMD Finances

I’ll keep the financial stuff brief.

Once you take the restructuring charge (which involves real money), and the income tax charge (which doesn’t; it’s basically just an accounting trick that will let AMD show bigger profits for a while in the future); the company lost $250 million.

It may seem funny to call such a result “good,” but AMD does appear to have reversed the downward spiral.

CPU revenues went from $250 to $420 million. That’s hardly a full recovery, but it does look like AMD is finally getting the Palomino off its back. Also keep in mind that AMD didn’t have a full quarter of relatively high-priced processors to sell last quarter.

Take those few factors together, and AMD should see another sizable jump in revenues next quarter, even if they sell fewer processors.

Are They OK Now?

Well, they aren’t on the express route to hell anymore, and should be able to muddle through the next couple quarters, but the second half of the year may prove to be a different story.

For the first half of 2003, AMD looks like it will follow a high-low pricing strategy. AMD will use one product line to cover two price points. Low-end TBreds have become AMD’s new Durons, and will be priced to compete against Celerons.

The high-end stuff will be priced high, and if that kills volume sales, so what? High-end TBreds (and shortly Bartons) are basically six-month stand-ins at the top end until Hammer arrives.

For instance, AMD said Bartons will be priced at around $300, which will probably mean somewhat less than that for the 2500+, somewhat more than that for the 3000+. That’s bad news for the impatient.

The good news for the patient is that the same processors will eventually and inevitably end up costing around $50, too (albeit a year or more from now).

The Next Peril: Competing Against Oneself

AMD is sort of like a corporate Indiana Jones right now. It looks like it got out of one predictament (not being able to get new Athlons out), but in the second half of the year, along comes an even bigger one: Hammer.

For the second half of 2003, AMD apparently plans to shove Hammer down consumers’ throats in one big shove. We’ll see these things showing up April, but production will be negligible: 5% or a bit less of total CPU production.

Normally speaking, phasing in a new generation of processor takes about a year. AMD apparently plans to do in a quarter, jumping Athlon64 production from 5% to around 50% in the third quarter, from around 300,000 processors to 4,000,000.

If they build it, will you come?

The overall plan is for Hammers to be the top-end product, and Athlons to assume second-tier status, but it’s hard to see how Hammer is going to put Athlons down that quickly. There’s just not going to be that much difference between the first generation of Hammers and the high-end Athlons/Bartons.

Pricing the two is going to be very tricky. If the “old stuff” is priced too low, everybody will buy that and not Hammers. If you keep the price of both old and new high, a lot of people won’t buy anything.

The core question really is: How flexible can AMD’s production be in meeting customer demand? Yes, they can use UMC as a foundry for Bartons, but how flexible can Dresden be in making or not making Hammers and making something else? If they make four million Hammers in the third quarter and only sell one or two, don’t we end up with a big inventory overhang again?

Perception will be important here. Saying you’re going to make between one-two million Hammers in the third quarter and selling them is one thing; that would be a normal rampup. Saying you’re going to sell four million and then selling one-two million is something else.

In the knee-jerk world we live in, Hammer could quickly be deemed a “flop” simply due to the hype surrounding it.

It does not help that Intel will have .09 processors out a few months after Hammer production gets serious, and AMD will have to follow a few months after that.

Finally, this is a weakened company with a lot more debt than it had a year; it’s like a video game character with its vitality points down. It can’t afford another strength-draining string of big losses.

That’s AMD’s next potential peril.

Email Ed