Tech Industry

Dirk Meyer, the man to watch at AMD

Company's president is being groomed to succeed CEO Hector Ruiz. But first he must prove that last year's engineering snafus were an aberration.

Imagine you're a director for a company that just endured a year that started off with a distribution snafu and then unraveled in a series of engineering glitches, forcing the delay of your most important product in years. Worst of all, your company's stock price dropped 63 percent.

It might be time to start thinking about going in a different direction, right? Not at Advanced Micro Devices. CEO Hector Ruiz said last month that he plans to stay on the job through 2008, and it doesn't appear that AMD's board of directors is inclined to send him packing.

Perhaps that's because they are still auditioning his replacement. AMD confirmed last month what had been evident for more than a year: President and Chief Operating Officer Dirk Meyer is the heir apparent to Ruiz. That strategy became clear years ago through a series of promotions, and Meyer is credited by some as playing a key role in the company's day-to-day operations for about a year.

Meyer is considered an engineer's engineer and was heavily involved in the design of two of the most noteworthy server processors to emerge during the last 20 years: Digital Equipment's Alpha in the 1990s, and AMD's Opteron, released in 2003. But some current and former colleagues of Meyer's say they aren't sure he's the right fit for the top spot.

Dirk Meyer Dirk Meyer, COO
and president,
AMD

While Meyer has considerable technical credibility, he is said to lack the softer sales-and-marketing touch that's crucial in the corner office. And given the debacle that was the Barcelona processor launch, some wonder whether Meyer--who ran the company on a day-to-day basis while AMD struggled to ship the quad-core chip--may not have the engineering clout he once did.

This year will tell whether he's also the tough competitor and ace salesman the chip outfit will need.

"This company is difficult to manage. It requires a CEO with a lot of energy to go up against (Intel). The CEO needs to be a dragon slayer, given the way we compete," said one source familiar with Meyer's management style. AMD declined to make Meyer or Ruiz available for an interview.

The next six months are a crucial period for AMD, Ruiz, and Meyer. Ruiz promised in December that the company will be profitable in the second half of 2008. That won't be easy. AMD lost more than $1.5 billion during the first three quarters of 2007, and analysts surveyed by Thomson One expect the company to lose 35 cents a share in the fourth quarter. At least that's headed in the right direction, compared with 2006 fourth-quarter loss of $1.05 per share.

AMD is accustomed to boom and bust cycles, and it has some work to do getting itself out of this latest mess. But choosing the right successor to Ruiz, 62, might actually be the company's most important task in 2008.

The man in waiting
Meyer, 46, has worked for AMD since 1995. He has risen steadily through the ranks of AMD's processor group, taking over the division in 2001 and becoming the de facto second in command at AMD in 2005, when the company decided to spin off its flash memory business, making the processor group its only business line. AMD formalized Meyer's status in January 2006, promoting him to president and chief operating officer and naming him to the board of directors in November 2007.

Placing Meyer on the board not only signals to customers and Wall Street that he is an anointed successor, but also changes the way he's viewed by fellow directors, said Jon Holman, who heads up executive recruiting firm The Holman Group. It's very different sitting among your fellow directors as a peer and laying out your strategy, as opposed to making the big presentation as an executive, he said.

Current and former colleagues, who did not wish to be identified, almost universally praise Meyer as a man with "integrity and honesty" who commands respect because of the accomplishments during his career. He's not prone to surrounding himself with yes-men and isn't afraid of being challenged.

But they also acknowledge that much of AMD's troubles during the past year can be placed at Meyer's feet. One source familiar with AMD's internal operations said that during the past 12 months, Meyer has taken on much more of the day-to-day management of AMD, while Ruiz has focused on key customer relationships and globe-trotting to promote AMD's interests, such as its 50x15 project to promote technology adoption in developing countries and its antitrust campaign against Intel. Unfortunately, that also means Meyer has been at the wheel during the Barcelona ordeal.

Barcelona is AMD's first quad-core server processor, based on the original Opteron design. The company decided to build a chip in which four Opteron processor cores were integrated onto a single piece of silicon, which is hard to do but pleased customers who wanted to protect their investments in the chipset and motherboard technology used in AMD's earlier designs, according to executives. AMD representatives repeatedly declined to comment on Meyer's role in Barcelona's development, but he's been the man in charge of the processor division since 2005.

Barcelona was originally supposed to arrive in mid-2007, but a steady stream of problems pushed most of the shipments well into 2008. AMD shipped Barcelona processors to select customers in the fourth quarter of 2007 after introducing the chip in September. However, after the first customers started getting their chips, a serious bug was uncovered that required the chip to be reworked before it could be sold to the general public. Performance is expected to suffer as a result.

Because of all this, lots of customers who were looking for all-out server performance in 2007 went with the competition. Intel's decision to build a quad-core processor by assembling two dual-core chips into a package--derided by AMD executives as an inelegant hack far beneath their "true quad-core" design--meant that Intel has enjoyed much more than a year with the quad-core server processor market to itself.

Intel has gone on to release the second-generation of its quad-core chips, and has a third iteration planned for later in 2008 called Nehalem that borrows many of the design techniques that made the original AMD Opteron such a hit, namely the integrated memory controller and the point-to-point interconnects that directly link cores to their neighbors. Those features allowed AMD's processor cores to enjoy fast connections to memory and to their fellow cores on a multicore chip, improving the overall performance of a chip.

However, AMD's combination of problems doomed its financial performance, especially after it was forced to severely discount its aging dual-core chips to compete against Intel's products. AMD promised financial analysts in December that it will break even by the second quarter, but they seem skeptical: Bank of America's Sumit Dhanda cut his rating on AMD to "sell" on Wednesday, citing an expected slowdown in the economy in the first half of the year and AMD's own internal troubles.

CEOs at other prominent technology companies, such as Motorola, Yahoo, and Dell have slinked off stage for performances that don't even come close to AMD's troubles--which have also included a distribution gaffe that alienated channel partners, the reorganization of its sales organization following the departure of sales chief Henri Richard, and a pending write-off of some portion of the $3 billion in goodwill associated with the October 2006 acquisition of ATI Technologies. But in AMD's case, it's not clear where the blame should be directed.

"It's hard to discern among the executives, since they present a unified front," said Ross Seymore, a financial analyst with Deutsche Bank.

Ruiz told CNBC Europe on the eve of its December financial analyst conference that he had no plans to step down in 2008, and sources familiar with AMD's succession planning say that the company's performance last year has not accelerated the timing for Meyer's ascension to the CEO spot--pinned at late 2008 or early 2009. Still, sources familiar with the board note that just like any other CEO, Ruiz serves at the pleasure of the board.

Often when a company president is named to the board of directors, that executive is named as CEO within six months to 12 months of joining the board--18 months at the latest, said Stephen Miles, managing partner of leadership consulting in the Americas for executive search firm Heidrick & Struggles.

There's a reason for that. The practice of naming a No. 2 executive to the board of directors is fairly uncommon, occurring in roughly less than 1 percent of companies in corporate America, Miles said. Intel and New York Life have used that strategy in recent years, but the No. 2 generally only gets to sit on the board once they become the No. 1.

"When a company is under siege, they know the No. 2 is likely to exit so this is a way to lock them in as a retention maneuver. It signals that a succession plan is in place and the clock is ticking," Miles said.

A Meyer administration
While AMD needs leadership as much as the next organization, it faces a number of unique hurdles.

Modern PC and server processor development is a two-company game, and Intel's staffing and marketing budget dwarfs AMD's resources; AMD must be nearly perfect just to keep pace. Intel can afford to throw away billions of dollars on peripheral products and acquisitions, tear up its technology roadmaps, and still remain profitable.

AMD doesn't have that luxury; when it places a bet, it needs to make the right call.

That also means that it has to take risks, and not all of them will pay off. The original Opteron design was a huge risk in 2002, but it paid off in spades. Paying billions of dollars more for ATI than it was worth? The lasting effects of that decision, in which Meyer played an important role, remain to be seen.

So, what kind of CEO would Meyer make? Despite the problems with Barcelona last year, Meyer's obvious strength is his technical background. But sources who have worked with Meyer have some concerns: the CEO of a chip company is not the lead architect on every design; he or she is a salesperson, a manager, a negotiator, a disciplinarian, and a visionary.

There is a big gap between Meyer's skills with customers and those of Ruiz, according to one source.

"Dirk is still close to the Excel spreadsheet. Technical skills can be acquired and trained, but it's the soft skills that are more difficult for people and some never get them," the source said. Ruiz has transformed AMD from an also-ran in the world of business technology to a consistent presence in data centers around the world. And while having a hot product such as Opteron certainly helped, Ruiz has forged regular and real relationships with AMD's top customers in order to make things stick.

An AMD representative declined to comment specifically on the source's comment, but noted that AMD's chip group--on the back of the Opteron processor--enjoyed a period of sales and marketing success while Meyer was in charge of the organization.

Meyer will also have to win over--or get rid of--employees brought into AMD by Ruiz who remain fiercely loyal to the man but might be partially responsible for AMD's troubles, according to another source. Clearly, a lot of people must have had a bad year for AMD to reach this point.

Meet the new boss?
Despite its troubles, AMD isn't going anywhere just yet: PC and server companies don't want to rely on a single source for x86 processors, and Intel executives know they can't really afford to bury AMD and open themselves up to the same kind of antitrust scrutiny at home that the company is facing in Europe and Asia.

But just as it did in 2002 before the launch of Opteron, when the company was losing truckloads of money, AMD finds itself in a precarious position. After squandering its chance to make a dent in the quad-core server market during 2007, the company will have to hold the fort in 2008 against strong products from Intel while waiting for its next generation of processors.

Fusion is AMD's project to integrate a CPU and a GPU onto a single chip, which could dramatically improve graphics performance if AMD can meet the engineering challenges. This is why AMD snapped up ATI and--just like Opteron did--it could vault AMD ahead of its larger competitor. But AMD doesn't expect to have Fusion chips ready until 2009 at the earliest.

Perhaps a more pragmatic Meyer is now starting to assert himself following the Barcelona experience. In late 2007, AMD changed its strategy for Fusion. It now plans to introduce Fusion chips for notebooks in 2009 that will be a combination of a new GPU design and an existing CPU design, rather than trying to build a completely new design based on the "Bulldozer" core outlined in 2007.

The Sandtiger processor--based on the Bulldozer cores--will now have to wait until 2010, an AMD representative confirmed. The representative declined to elaborate on who was responsible for that decision, but noted that this time around, the company's customers approve of the decision to minimize risk.

According to sources, AMD's board has shown some irritation at the lack of AMD's execution, but a number of the members don't have a technical background in the PC and server industries. As a result, they rely heavily, like a number of corporate boards, on management's assessment and recommendations. Those recommendations haven't worked out so well over the past 12 months, and further missteps could be impossible to ignore. AMD declined to make any of its directors available for comment.

If the board of directors forces Ruiz to step down before Meyer is ready to take over, AMD's struggles could continue. And perhaps more troubling is the fact that if AMD continues to struggle under Meyer's guidance in 2008, the company may have to shred its succession plan and start from scratch.