It's No Rumor: AMD CEO Dirk Meyer Resigns 18 Months In
What makes this situation unusual are all the things AMD doesn't say. The press release neither gives reason for Meyer's departure nor explains why Dirk is being ousted immediately after launching the Fusion APUs first envisioned by his predecessor. The board's statements, however, do hint at a potential disagreement between the former CEO and the rest of his colleagues. Immediately after praising him for his achievements above, the document states:
However, the Board believes we have the opportunity to create increased shareholder value over time. This will require the company to have significant growth, establish market leadership and generate superior financial returns. We believe a change in leadership at this time will accelerate the company's ability to accomplish these objectives.An Odd Choice of Words
To explain the strangeness of that statement we have to delve into the past. Meyer came to AMD in 1996 along with a team of engineers from the dying company DEC. Meyer led the K7 design team and was therefore credited for the processor that literally saved the beleaguered manufacturer from collapsing under Intel's weight. The company promoted Meyer to head of the Computation Products Group in 2001, made him a senior vice president in 2002, and appointed him its Chief Operating Officer (COO) from 2006-2008.
Here's where it all began—AMD's original Slot A K7 Athlon, built on a 180nm process
When Meyer finally took over as CEO, he replaced Hector Ruiz. Ruiz had guided AMD through some of its greatest successed but had also officated some of its greatest missteps and failures. It was Ruiz who masterminded the plan to spin AMD's manufacturing off as an indepedent business, but the older CEO was swept out of office after Sunnyvale posted its seventh straight quarterly loss. When Meyer ascended, he was painted as the perfect candidate to clean up Hector's mistakes, put the company back in the black, and finish the much-delayed Bulldozer and Bobcat. As far as we know, that's exactly what he's done.
After 14 years of positive perception, 18 difficult (but successful) months as CEO, and with AMD's radical new architectures just half launched, Meyer—the chief architect behind the CPU whose basic design has powered every AMD processor from 1999 through 2010—is out without even a cursory explanation.
The Dubious Merits of Increasing Shareholder Value:
It's possible that Meyer's departure was caused by events of which we are currently unaware, but the BoD's earlier statement implies that it and Meyer profoundly disagreed on AMD's course. If we had to guess, we'd bet that the board, already frustrated over AMD's inability to capitalize on the netbook craze from a few years back, is similarly peeved that it didn't have tablet products ramping over the last six months. While netbook/notebook Fusion parts are already shipping, tablet versions of those processors won't be ready until the end of Q1; we may not see AMD push into the tablet market until back-to-school 2011.
Here's where we are now, but Zacate and Ontario still have to prove themselves in a crowded, Intel-dominated market
The problem we have with our own prestigious prognostications and clairvoyant cogitations is that they don't make much sense. Tossing Meyer because AMD didn't have netbook/tablet class processors available earlier or for not aggressively planning to ramp tablets now would be an enormous mistake. The initial netbook explosion took even Intel by surprise and, as we've commented before, was only possible because Santa Clara had specifically designed a new processor from the ground up. AMD had nothing that could match Atom's power consumption and sensibly opted not to try. The fact that the netbook market caught like a magnesium flare only to shrink dramatically in 2010 (again catching Intel offguard) is further evidence that Meyer made the right choice when he opted to play the interim period from 2008 to the present day slowly and carefully.
AMD made a critical mistake in 2005 by assuming that flattening Prescott and the Netburst architecture was the same as flattening Intel. While Sunnyvale chased ATI and brokered what was possibly the worst deal in its own history, Santa Clara was busy with the first iteration of the Core architecture. AMD's board makes noise about improving shareholder value, generating superior financial returns, and establishing market leadership. It sounds good, but it's ominously familiar and out of synch with the reality of where the company stands today.
What AMD needs to focus on is staying solvent and launching Bulldozer and Llano while carefully architecting its next-generation CPUs and GPUs. Hopefully Meyer's replacement realizes that as well. If he doesn't, things could get very ugly, very fast.