At the moment, AMD seems to be having two problems:
Big Bad Blue: This shouldn’t be any news to anyone reading this website. Intel pounded AMD with its “when you have lemons, make lemonade” policy with its Pentium pricing, now they’re going to pound them with cheap C2Ds. The purported April 9 price cuts looks like an appropriate response to Intel’s late April cuts.
AMD’s current problem at the moment is that Intel isn’t its only problem. The other problem is:
Dropping the Ball: Over the past few weeks and months, we’ve heard, in dribs and drabs:
All are probably true to at least some degree, and this is what is causing these price cuts every few weeks.
That’s bad enough, but when you have a growing number of people, including financial analysts, now focused on any sign of weakness, well, shooting yourself in the foot is not the best response to being bombarded by the enemy. The whole becomes more than the sum of its parts.
What’s Going To Happen?
AMD will probably lose a couple hundred million dollars this quarter, and do a bit worse than that next quarter. That may look better than last quarter, but the five hundred million dollar loss was almost entirely due to a big write-off. These losses will be due to current operations.
AMD will have to get more money within the next six months. No matter how they do it, they will have to get it under less than ideal terms.
What kind of price will AMD have to pay? The real price they’ll have to pay is a set of clipped wings. The “do or die” strategy will have to change into “entrench and hold on to most of your gains.”
What is that likely to look like?
Unless AMD gets bought out by somebody who can buy the company and then inject a few extra billion dollars into it (which pretty much excludes private equity firms), this is will probably happen by the end of the year:
The financial people would like AMD to cut its capital spending, but at a minimum, AMD has to renovate Fab30 to modern standards, for technical, capacity and political reasons.
What AMD doesn’t have to do is build a third fab in New York State. That idea is going to die, or at least be put on a very small back burner until much happier days are here again.
The price war stops with the next generation of processors, and if that means giving up a few points of marketshare, so be it. Forget 30% or more anytime soon, keeping 20-25% will become the real goal for the next few years. AMD can try again in a couple years with Fusion processors, after they’ve made some money for a while.
In short, “Do or die” will be replaced by “Hang on and make do with Dresden,” and if Hector can’t swallow that, Hector will no longer be running the company.
I’ll point out that if these rumors are true and AMD is bought out by private equity, they’ll probably do exactly the same thing, and chuck Hector out quicker.
You see, what was bad about “do or die” wasn’t “do or die” in and of itself. What was bad was trying to do it without money (sitting on their Hammer laurels thinking Intel couldn’t answer it, but again, that boils down to lack of money). Add five billion dollars in AMD’s piggy bank, and “Do or die” is a fine strategy likely to succeed.
What wasn’t so good was substituting five billion extra dollars with acting like you had an extra five billion dollars, hoping to make it by faking it. That worked for a while, and fooled just about everybody except the one that ultimately matters: Intel. They took some shots until they were in position to show that the Emperor Hector had no financial clothes.
The irony of all this was that AMD could have gotten that extra five billion, or even more a year ago when the stock was riding high. There would have been some difficulties, but odds are Intel would have thought twice about a price war knowing that AMD had the cash to ride one out.
But rather than bolstering the piggybank, AMD emptied it buying ATI, leaving it financially vulnerable just when the Empire was striking back.
Sad, but it just illustrates what has always been AMD’s core inherent problem: lack of money compared to its competitor.