Ir doesn’t make sense. It just doesn’t make sense.
AMD spent tons of time saying what great shape they really are in, and that they’re primed and ready to go for Christmas.
Unfortunately, out of all the numbers they were throwing around, they tossed out one that made the rest of what they were saying sound like the captain of the Titanic radioing to shore, “What a great cruise! Free ice and free delivery!”
AMD’s revenues were a little bit over $500 million. That broke down CPU: $262 million; Flash: $189 million; other: $57 million.
AMD said they expected revenues for the Christmas to be up 20%, or about $100 million dollars. That sounds pretty good until you realize AMD’s CPU revenue are down over 60% from about $700 million six months ago.
Let’s take AMD at their word, and presume they’re correct.
AMD said they expected flash to do well next quarter, so some of that $100 million increase will go there. Flash sales actually were up 8% the previous quarter; let’s presume another 8% increase even though it will probably be more than that. We’re doing this to try to come up with the highest somewhat realistic CPU revenue estimate.
That leaves about $85 million in increased revenues. Let’s put all of it on the CPU side.
$262 million + $85 million = $347 million dollars.
In other words, about half the revenues of Q1, and 10% less revenue Christmas quarter than the horrible Q2, in which AMD sold about six million CPUs for an average of little over $60 each.
But wait a minute. AMD is busy telling me those horrible sales figures for Q2 and Q3 were due to cutting back production to lower inventory levels, and they said they’ve pretty much taken care of that.
Wait another minute. Unlike Q2, AMD now is going to have all these new processors with much higher price tags than the old Palominos did six months ago. And they said things that sounded like their 130nm process was fine, and they “can make” two million 2400+s this quarter.
So if you can sell more new processors for more money in a quarter that should have more demand than six months ago, how can you possibly expect to get less revenue than six months ago????
Missing: A Hundred Mil or Two
If everything were honky-dory at AMD, given all the other cirumstances, even if you’re being conservative, you don’t come up with a number like $350 million. You come up with a figure more like $450 million, and if you have any kind of optimism about AMD at all, you come up with a number a good deal higher than that.
Mind you, these are still pretty awful numbers compared to six months ago, and I’m sorry, but AMD selling eight million processors in Q1 was no freak accident; they had been selling around that number for at least a year prior to that.
AMD did not provide the number of processors sold last quarter, and refused to provide it when asked. All we know from the press release was that it was less than last quarter’s six million, and the execs estimated Christmas volume would be only 8-12% better than the Unknown Sales-or. They essentially said, “Try to figure it out like you have to for Intel.”
Nor is the CPU market in any great decline worldwide; it’s basically static. Intel may have dropped a couple percentage points in CPU revenue, not 60%.
With all due respect to Intel, not only is it hard to believe that they’ve whipped AMD’s butt that thoroughly, but it’s infinitely harder to believe that they have AMD executives so thoroughly convinced of that, too.
Let me put it another way. No one comes here for my raving optimism, and certainly nobody will accuse me of that about AMD the last few months.
In contrast, I think all would agree that Chris Tom of AMDZone views AMD somewhat more positively.
Hearing those AMD execs giving me happy news, then coming up with much lower Christmas estimates than grumpy Ed is like me coming up with a gloomy AMD revenue estimate, and Chris Tom coming up an estimate 25% lower than mine. You certainly can’t say he’s biased against AMD. The only logical conclusion is that he knows something I don’t, and it’s bad.
Some might say that AMD is deliberately low-balling the figure so as to easily exceed it. That’s an understandable reaction, but when you’re trying to convince a bunch of analysts and investors that your company is not headed down the tubes, grossly understating the one measure that would reassure them is a very strange way to go about it.
No, it makes more sense to believe that those AMD execs know something bad we don’t, and they’re not telling us what it is.
And that’s bad.