With a fresh playbook, AMD's Dirk Meyer is shedding costly chipmaking operations to focus on semiconductor design, restoring profits, and beating giant Intel
You might say Dirk Meyer just had a great big weight lifted off his shoulders but still has another to bear. As the new CEO of chipmaker Advanced Micro Devices (AMD), he has taken the helm of the company just as it has closed a complex spin-off of its manufacturing assets into a new company called Globalfoundries. The deal is the second major change to AMD in recent years. In 2006 it paid $5.4 billion to acquire the Canadian PC graphics chipmaker ATI Technologies, an acquisition that proved ill-timed. AMD ultimately wrote down the value of ATI’s assets by half.
While AMD's chips have at various times been competitive—and occasionally superior—to those of its chief rival, Intel (INTC), the company has always struggled with the costs associated with building and operating multibillion-dollar chip factories, known in the industry as "fabs." Last year it lost $1.95 billion on sales of $5.8 billion.
The new Globalfoundries, to be half-owned by AMD and Abu Dhabi's Advanced Technology Investment, will build AMD's chips under contract but also will seek chip manufacturing business from other companies. There are already several so-called chip foundries, among them Taiwan Semiconductor Manufacturing (TSM), United Microelectronics (UMC), Singapore's Chartered Semiconductor Manufacturing (CHRT), and even IBM (IBM).
Now that the crushing capital expense of manufacturing is off AMD's books, Meyer still has a big job ahead of him: taking on Intel, one of the most powerful companies in technology. He sat down with BusinessWeek's Arik Hesseldahl to discuss the state of the microprocessor market, the PC industry, and AMD's new structure.
Dirk, the biggest change with this new structure, at least looking from the outside, is that AMD won't be in full control of its factories. How is this new operating structure going to be different in an operational sense from what you've been doing all these years?
On the margin I would say it's a little different, but really it's not all that different. We're still going to have to provide Globalfoundries with good forecasts. And we're its only customer coming in and will be its biggest customer for quite a long time, so they're going to be very motivated to give us what we need.
Are there likely to be any other customers coming in?
They've got a lot of good discussions going on. One of the ideas behind this deal is that since R&D and [capital expenditures] have been growing faster than industry growth rates, you're seeing product companies that aren’t able to scale for manufacturing. Therefore the manufacturing base of the industry is consolidating around companies that do have the scale, and most of those companies have turned out to be foundries. So we observed that when you look at the foundry players today, you've got TSMC [Taiwan Semiconductor Manufacturing], which is the only one with technology at the leading edge and a strong balance sheet. The rest of them frankly have neither. We saw two things: First, among the companies looking to hire a foundry today, they're looking for an answer other than TSMC. Second, we saw an industry that was increasingly going fabless. So for both of those reasons, we saw an opportunity to start a foundry company.
What's the most important thing for AMD now that you're out of the manufacturing business?
We've got a good intellectual property portfolio in the form of PC and server CPUs, graphics, and video. We're focused on being a bit more agile in the way that other nonmanufacturing companies are. The culture of a company is different when you've got these wafer fabs. It's just a different mindset. The planning horizons are longer than when you consider yourself by culture and necessity a design company. So that is the kind of mindset shift that's taking place now.
What do you think of the reports that Nvidia (NVDA) is interested in making x86-based chips like you and Intel do now?
I'm not surprised by it. It confirms the idea we had when we bought ATI [Technologies] a few years ago. I think it's pretty ironic that Jen-Hsun Huang, who's a good friend of mine, was touting the fact that Nvidia was the only independent graphics chip company left as if it were a good thing. In my view it's not a good thing. But I'm a little suspicious of Nvidia's ability to create an x86 chip from scratch. The intellectual property barriers alone are enormous. It's not something you switch on like a light switch. And by the way, you have to pay [Intel] for it.
So the latest market forecast numbers from Gartner (IT) show the PC market shrinking by 12%, while IDC says it will shrink about 5%. What's your view right now? And with the market contracting, what are your priorities?
The Gartner numbers are worse than what we're seeing right now. We're thinking more single-digit unit declines, which would be the first time since 2001 and only the second time in the history of the industry, so that's not good. The mindset is now all about cash management. We also want to make sure we maintain the investments on R&D so when the recovery occurs we've got the right mix of products at the right time.
What does the new structure as a fabless company do for the prospects of profitability?
We've restructured the company to bring the break-even point down from roughly $2 billion in revenue a quarter in 2007 to $1.5 billion a quarter. In light of the change in the environment, we brought that break-even target down even further, to about $1.3 billion. The other thing that's very popular with this deal is that it changes the cash-flow profile of the company. Historically we have had a negative cash flow even when we were making money at the net income level because of the huge investment in [capital expenditure]. Going forward we see positive free cash flow at the $1.3 billion level, which is a very different model from what AMD has done in the past.
What's in the pipeline right now?
First, on the server side we started shipping 45-nanometer quad-core parts last quarter. The next thing will be a six-core upgrade that will plug into the same sockets as the quad core. On the client or PC side, we've got a bunch of new introductions queued up starting up this year and going into next year on graphics, notebooks, and desktops for the back half of the year. After next year, we'll have two classes of client [products]. Habitually we and Intel build chips for the mainstream and high end of the market. We'll still have those development tracks, but on the PC side we're emphasizing notebooks. We’ve also opened up a development track aimed at lower-cost and lower-power consumption machines, kind of in the spirit of netbooks, though I think the netbook name will go away over time.
Why do you say that?
Well, what's a netbook? It's a small notebook. But how small? Where's the cutoff point between a notebook and a netbook? I think the line will continue to blur and people will just think of them both as computers.
Yet that is one PC segment that's a little healthier right now and AMD isn't participating in it. Why is that?
We consciously focused our R&D dollars, which obviously given our size are smaller than Intel's, on the big mainstream markets as they exist today. Knowing that this trend toward lower power consumption and more mobility is going to happen, we just decided to load that into the R&D pipeline for later. It's not a big volume target and not a big dollar opportunity.…One of the saddest things about the PC industry right now is, since late last year, all anyone seems to want to ask about is netbooks. Good grief! It's a low-cost limited-function device. There's not much excitement or money in dollar volume there.
So what do the PC makers tell you about their outlook for right now?
No one is expecting anything miraculous in the way of a turnaround in 2009. And we've said publicly we're not, either. If it starts to turn around I think that's great, but who knows. I was just in Chicago visiting enterprise customers, and if you look at the overall IT spending, I'd say key IT investments are still happening. We haven't found anyone who plans to spend more this year vs. last year. Most plan to spend a little bit less. But people are still saying they have to invest for the future, and IT still drives productivity and lowers costs. To the outside observer it must look like a complete and total meltdown, but it's not like automobiles—knock on wood.
How do you expect to do things differently vs. your predecessors Hector Ruiz or and Jerry Sanders? How are you different?
One thing we all share is that we all want to win. That's one thing I talk about when I walk around the company. Why are people at AMD? It's a tough business. It's a small guy going up against the biggest guy in the industry. It's because you want to win and change the industry. Imagine a world where the world's most important information technology only comes from one place. Nobody wants to live in that world. Jerry introduced the theme, Hector picked up that theme, and I intend to continue that theme.