CPU 2117

AMD had a bad first quarter. They went from a $160 million loss in Q4 to a $360 million loss in Q1. While an increased loss was expected, this was somewhat bigger than expected.

90% of that increased loss was due to the CPU business. This was because people (most notably Dell) got increasingly tired of buying old stuff, and AMD made hardly any new stuff. So they sold roughly 15% fewer CPUs last quarter, as opposed to a “normal” seasonal drop of 5-10%.

You might say, “Well, won’t things get better next quarter when AMD will have lots of new stuff available?” Uhhh, no. I’m not predicting that, AMD is. They expect revenues to drop a bit more next quarter, another 5-10%.

It doesn’t take a rocket scientist to figure out that AMD isn’t going to make all too much new stuff next quarter, and what extra money they might get from the new stuff will be more than offset by fewer sales/lower prices on the old stuff.

Combine that with an extra cost for laying off 10% of the staff, and AMD will lose more money next quarter, probably somewhere in the range of $400 million.

Matters will probably improve somewhat later in the year, but again, not even the AMD execs are promising a true breakeven any time soon. 45nm chips are now supposed to show up in Q4, but even if they’re good, the production numbers will be small.

Since AMD has no prospects of making any money soon, the tactics for the next few months appears to be 1) reducing costs and 2) selling assets.

Layoffs are one way to reduce costs, though anticipated savings are only expected to be $250 million a year, less severance charges. That’s only a small fraction of what’s needed.

They practically put the consumer electronics unit from the former ATI up for sale at the conference call. Unfortunately, given its experiences with buying ATI and selling Spansion, AMD has the bad habit of buying high and selling low.

Non-fab properties are available for sale/leaseback, and AMD is dropping bigger and bigger hints that it would like to do something like that with its fabs, too.

However, while sale/leaseback can get you short-term cash, in the long-term, it raises AMD’s costs because now they have to pay rent on the properties.

At least sale/leaseback on the non-fab properties shouldn’t be too hard to arrange when needed. The fabs appear to be a different story. AMD’s “asset lite/asset smart” strategy appears to be “find somebody who’ll buy a minority stake in our fabs.” It also appears that despite a year of looking, they haven’t found any takers.

We could go on and on but one fact has become blindingly obvious: this company can bumble along hand-to-mouth for a while, but eventually, they’re going to have to be bailed out, one way or another. This company is not going to be able to recover on its own. Period.

There are lots and lots of ways AMD could be bailed out. How likely is this to happen? We don’t have a clue, and it’s not just a matter of AMD secrecy, I doubt even Hector has much more of one. The long-term fate of AMD is out of AMD’s hands: it’s going to be in the hands of judges, politicians, financiers and economic conditions, most likely in some wildly variating combination of all the above.

For the meantime, one can only say the following:

  • Anyone who says AMD is sure to recover is a fool. This company could die.
  • Anyone who says AMD is sure to die is just as much a fool. This company could live.
  • If a decisive thumbs up or down happens, it will happen quickly, and we will almost certainly be blindsided by it.


  • Be the first to comment

    Leave a Reply