AMD: 2008 . . .

Add Your Comments

Yesterday, we covered the major financial points made at the analyst meeting. We then put high finance on the backburner since the company isn’t going to go under anytime soon, which is all most reading this are concerned about.

However, the analyst meeting also revealed that the financial crisis is going to have a big impact on AMD’s future development plans. This will affect all of us, financially inclined or not.

New announcements, combined with some earlier revelations, indicate that AMD is going to run the near-future on the cheap, and put anything else not already deep into the pipeline on hold until better days.

The plan seems to be to spend as little as possible for the next year or so, make do with whatever resources they have, get some modest successes, make some money and make Wall St. happy, then resume business as usual.

AMD’s strategy the last few years has been “Do or die.” Wall Street’s response is now “Die,” so the strategy shifts to “I don’t want to die yet!”

Can the new strategy succeed? Let’s walk through AMD’s plans for 2008:

K10s: For better or for worse, the production K10s will show up around March. They won’t be as fast as Penryns, or even the faster C2Ds, but that’s spilled milk at this point.

The big issue facing AMD with K10s, particularly Phenoms, is quite simple. Can AMD make significantly more money making them than they can with the current K8s?

For K10s to give AMD financial help, a lot of them need to be made, and they need to be sold at a rather higher price than K10s.

Let’s look at these two issues. It doesn’t look like Phenoms will ramp up in a flash quickly. You may see the first ones in February or March, but it probably won’t be until the summer that the K10 generation will become a substantial part of AMD production, and the fall or later that it might account for the majority of production.

That’s not a slow ramp, but let’s face it, these guys are under the gun. On the other hand, they don’t have the money to throw at this problem to get a faster ramp, either.

More importantly, AMD has to get more money for its new CPUs than it gets for its current ones. They can’t keep getting less than $60 average per desktop and mobile chip from the OEMs like they are now. That’s going to be difficult to change.

Intel will certainly make as much pricing mischief as possible by lowballing the lower-end CPUs to try to push Phenoms into the bargain basement, but AMD’s biggest foe will be themselves with their K8s.

Since the quadcore market is unlikely to become mainstream in 2008, only a small portion of AMD’s K10s are likely to be quads once serious quantities start being made. The bulk of production will likely be duallies, but K10 duallies just aren’t going to be much better than K8 duallies. If a 2.2GHz K10 performs the same as a 2.5GHz K8, how is AMD going to get more money for the K10s from Dell?

Looking at quads, well sure, AMD will get more for a quad than a dually, but how much more? If AMD ends up getting less money for a quad than it does for two duallies, that doesn’t do them much good, does it? If AMD has to break a quad to sell it for less as a tri, that’s even worse, isn’t it?

This isn’t to say AMD is doomed, but this is going to be a tough hand to play with the cards they have in their hands.

45nm Some believe AMD needs to convert to 45nm as quickly as possible to survive, but there’s problems here, too.

The slides at the analyst show imply/say that AMD will start making production 45nm chips at the end of June and have them ready for sale in September/October. That isn’t going to ramp up very quickly, either, 45nm won’t hit mainstream chips until around mid-2009.

It’s OK to wonder about AMD being able to meet these targets, given what happened with 65nm. AMD apparently does, too, but their solution to the problem may end up being worse than the disease.

AMD’s 45nm chips are not going to be made using IBM’s latest and greatest technologies, using high-K dielectrics and metal gates. Rather, they’re going to stick with what they’ve been dealing with, adding immersion technologies to be able to work at 45nm.

This is a cautious approach to 45nm, one being done on the cheap. You have to wonder, wonder a lot, whether AMD went off on a different path due to lack of money.

On the one hand, this approach probably maximizes the chances AMD can get something at 45nm out the door more-or-less on time. On the other hand, it also maximizes the chances the end results, even if everything goes right, will be even more lackluster versus Intel’s Nehalems than K10s will be against Penryns.

They would be cheaper to make, though. Is that enough to offset the lower prices from lower performance?

The Future Ends Now Maybe AMD got advance word the world would end in 2008 because they seem to have scrubbed most of their post-2008 plans.

On the ATI side, R700 won’t be out until 2009. The new CPU designs promised in 2009 have gone MIA, initiatives like Fusion will just reuse 2008 parts. Outside of switching to DDR3, mobos look the same way.

Sure looks like they’re abandoning the future to stay alive next year, and even if this meant to be just a tactical retreat until a day when they can get more money and be bold again; that may be as much wishful thinking as the old “Do or Die.”

It’s not much of an exaggeration to call the thinking behind this, “We’ll get better prices for worse products.” Even if you more accurately call it, “We’ll get the same, but it will cost us less to make,” that’s not a viable long-term strategy, either.

To be fair, it’s not like the top people at AMD think this is a great idea; it’s much more like least bad choice. These people aren’t that stupid; brave words notwithstanding, they know they’re in deep, deep doo-doo among the financial people, otherwise they’d never do this.

The question becomes, “Can AMD make money putting out deliberately second-rate products against Intel by saving on R&D and new technologies?”

And frankly, for the long-term, and maybe even in the short term, that sounds like desperate actions being taken to delay the probable death spiral. That’s not the intention, but unless fortune smiles on them more than it has lately, that’s a pretty likely result.

Ed


Leave a Reply

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