AMD Financial Results

AMD lost money last quarter, but the CPU part of the business wasn’t responsible for that. They lost some money in flash memory, and, to make a long story short, it cost them $50 million to turn a lot of short-term debt into long-term debt. That was something they needed to do eventually, anyway, so showing a loss this quarter is no big deal in the long run.

Their performance in the CPU industry was a bit odd, though. They increased unit sales of desktop Hammer 70%, and half of (about $700 million) CPU revenues are now coming from Hammer (as previously predicted), but yet revenues only went up 9%, and they didn’t make an extra dime from all the extra effort. They made more CPUs (they said unit sales went up over 15% to an all time high, which means around 8.3-8.4 million) but ASPs went down.

There’s a few reasons for that. First, they sold more cheap Semprons, especially in places like India and China, than they had expected. Second, average Hammer prices went down. Third, marketing of CPU products were sharply up for the Christmas quarter.

I don’t know. Seems to me that so long as AMD has only Dresden, all those cheap Indian and Chinese Semprons are going to be an anchor on revenues, ASP and profits. and maybe a few other things, too.

What They Didn’t Say

When you listen to an AMD conference call, what they don’t say is often more important than what they do say.

In this conference call, what AMD didn’t talk about was upcoming technological advances. No mention of strained silicon. No mention of improved clock speeds. They did mention that Turion ought to show up around May. DDR2 came up, and while AMD was non-specific, it sure didn’t sound like it was coming anytime soon. Since the flash business was “dismal” (Hector Ruiz’s words, not mine), AMD was trying to talk up the CPU side of the family as much as possible, the omission is glaring; you’d think they’d mention it in passing.

What They Did Say

The only other item you might call a “technological advance” is 90nm conversion. AMD said its starts are around 50% 90nm now, and should be about 100% (excluding legacy products) by Q2.

This raises a few questions. First, does that mean just about all Hammers will be 90nm by Q2? That would be good news to current or prospective socket 754 owners.

A bigger question is: What happens to all those cheap Indian and Chinese socket A Semprons? Might they go 90nm too? No matter what happens with Hammer, odds are there will still be a market for a few million of those for at least the rest of the year. They’ll certainly be cheaper to make than 130nm chips. Does AMD leave a big chunk of Dresden at 130nm (limiting Hammer production) while they try to phase out XPs in favor of 90nm Semprons, or do they make a few 90nm XPs, too?

I could be very well wrong on this one, but my gut tells me there’s going to be some 90nm socket A chips coming out in the next six-eight months just like there were 180nm K6-2s a few years back and 130nm Durons more recently. Probably not a lot, and they will be aimed at the developing world, but I’m sure a few will leak elsewhere.

Another item of interest is that Hector has changed his tune about market share (before, he used to say that pricing was more important than market share), and now thinks AMD ought to aggressively try to get more of it.

Well, he’s not going to make much money selling socket As to India/China, no matter how many he makes, 90nm or not, and AMD promised to aggressively cut back on its Hammer-related marketing expenses, so I don’t see how he’s going to do that without cutting prices on Hammers (though that could just mean sub-$100 Semprons).

Well, there is now one other possibility. 🙂


Warning to the humor-challenged: The following is not at all serious.

While talking about the quarter’s flash memory losses, Mr. Ruiz made the quite-unusual-for-him remark that, “It makes me puke to lose $39 million.”

What a motivational device (especially if you’re in his immediate vicinity)! Who needs “You’re fired!” when you can do this instead?

I can just hear the other AMD execs nervously whisper to each other, “I heard that when you lose more than $50 million, he loses control of his bowels, too.”

Imagine what he was doing when AMD was losing $200 million. He must have been a fountain of orificial outflow. I bet that got the other AMD execs on board for downsizing fast when he took over. I’d fire my mother after seeing something like that.

I can just see the business books coming out extolling the Bad Burrito Boss style of management. Why not? It’s as American as apple pie, and a proven success at the very highest levels.

Thirteen years ago, the President of the United States went to Japan for trade talks. For decades, America tried to get Japan to open up its trade, to virtually no effect.

Clearly, it was time for a new approach. Let the facts speak for themselves:

1/7/1992: President Bush arrived in Japan on a tough-talk trade mission.

1/8/1992: Bush vomited on the Japanese prime minister’s lap.

1/9/1992: President Bush declared his trade visit to Japan a success, saying Japanese officials had agreed to increase imports of American cars, auto parts, computers and other goods.

And really, economically, Japan hasn’t been quite the same since.

Need I say more?

Now that we’ve established how persuasive puking can be, maybe that’s what AMD’s new “aggressive approach” to increasing market share is going to be. Dell won’t buy? Just send “Regurgitatin’ Ruiz” in there mano-a-mano with Mikey Dell. Do a little ‘gurgin’ with Bill Gates, and see how fast Windows for x86-64 comes out.

Now imagine grassroots ‘gurgin’. You’re a salesperson at CompUSA and somebody wants a mini Mac or Centrino notebook rather than an AMD product you’ll get twice the commission on. Spare the idle chatter. Just go, “Mini Mac? Bleeaaahhh!!!” That will have an impact.

In an age when any publicity is considered good publicity, the time has come for barfketing. It’s aggressive; it’s cheap, it’s easy, and boy, does it makes an impression!



Be the first to comment

Leave a Reply