AMD: Too Needed to Fail

AMD continues to suffer through corporate misery, most recently by losing almost $1.2  billion in a single quarter, forcing the replacement of CEO Hector Ruiz with his subordinate, Dirk Meyer. If AMD collapses and Intel becomes the only major vendor of PC processors, will prices soar?

Unfortunately, monopolies usually do inflate prices. They also retard progress. AMD stimulates Intel to price its processors more aggressively and develop better processors. Without AMD, we might not have 64-bit x86 processors today or PC processors with integrated memory controllers. Ri

ght now, we’d probably be looking forward to the first quad-core x86 processors instead of the first eight-core chips.

Behind AMD, there isn’t much competition. In third place is VIA Technologies, a Taiwanese company with a small x86 design team in Texas. VIA doesn’t challenge Intel’s high-performance PC and server processors, as AMD does. Instead, VIA sells low-power x86 chips. But even at the bottom of the barrel, VIA scrapes for a measly 2 percent market share, and Intel’s new Atom is surprisingly stiff competition.

Too bad AMD isn’t a big Wall Street bank. Then the Federal Reserve could bail it out.

Actually, a government rescue would make more sense for AMD than it does for most banks. There are thousands of banks, but only two large companies making PC processors. Overnight, AMD’s demise would create a monopoly that’s almost impossible for another company to break.

Oddly, monopoly status could harm Intel’s business plans. Contrary to popular belief, U.S. law doesn’t forbid monopolies—it merely forbids companies from using monopoly power to dominate other markets. If Intel gains a monopoly in PC processors, any move toward dominance in other processor markets might be challenged in court. Intel has enough legal problems already with lawsuits alleging anticompetitive practices. Monopoly status could prevent Intel from diversifying in important ways.

Therefore, even Intel has reasons to keep AMD alive. PC users do, too. AMD isn’t “too big to fail,” but it might be too needed to fail.

Tom Halfhill was formerly a senior editor for Byte magazine and is now an analyst for Microprocessor Report.