FTC Sued Intel, nVidia dreams x86

source: nytimes.com
F.T.C. Says Intel Stifles Competition
By STEVE LOHR

The muscular giants of the technology industry often seem to be one step ahead of antitrust regulators, who must rely on century-old laws and slow-moving courts. Technology markets, by contrast, operate according to the hurry-up clock of the Internet age.

The Federal Trade Commission, one of two main agencies that enforce the antitrust laws in the United States, is trying to pick up the pace — and look to the future.

In an unusual complaint filed Wednesday against Intel, the world’s largest chip maker, the F.T.C. said it wanted to address anticompetitive abuses not only in the past and present, but also down the road.

Armed with a hybrid law that spans consumer protection and antitrust concerns, the agency said it was seeking to prevent Intel from using its dominance in the market for microprocessors, the main control chips in personal computers, to squelch competition in video graphics chips. The graphics-chip market is currently quite competitive, with Intel facing off against Nvidia and Advanced Micro Devices.

In its complaint, the F.T.C. said that the statute gives the commission “a unique role in determining what constitutes unfair methods of competition” and the power “to stop in their incipiency” acts that, if allowed to continue, would violate America’s two primary antitrust laws, the Sherman Act and the Clayton Act.

“The agency is very focused on the danger that these high-technology markets move on,” said George S. Cary, a partner at Cleary Gottlieb Steen & Hamilton in Washington, who is a former senior F.T.C. official. “They are looking to the future.”

In its complaint, the F.T.C. accused Intel of conducting a systematic campaign to block rivals from selling their microchips by cutting off access to the market.

The filing goes beyond charges in cases brought recently by European regulators and the New York State attorney general, which focused on microprocessors. Intel supplies about 80 percent of the PC microprocessor chips used worldwide.

The F.T.C. move, according to Andrew I. Gavil, a law professor at Howard University, is “very significant because it is broader in scope than any of the current cases.”

The agency contends that Intel engaged in a pattern of conduct that “put the brakes on superior competitive products that threatened” its microchip market share.

“Intel has engaged in a deliberate campaign to hamstring competitive threats to its monopoly,” said Richard A. Feinstein, director of the F.T.C.’s Bureau of Competition. “It’s been running roughshod over the principles of fair play and the laws protecting competition on the merits.”

The agency’s action comes after a yearlong investigation — and one month after Intel reached a sweeping $1.25 billion settlement with its longtime rival in the chip market, A.M.D.

That settlement, covering both private antitrust and patent claims, was seen as possibly deterring the agency from moving ahead. In its long-running legal fight with Intel, A.M.D. was both the principal victim of the giant chip maker and its nemesis, generating most of the evidence that was then used by government regulators around the world.

Intel and the F.T.C. tried to reach a settlement, but those talks foundered in recent days. Intel contends that the agency’s staff brought up the accusations about graphics chips and chip-related software late in the talks, asking about some details as recently as Dec. 8.

The F.T.C. action was “misguided and unwarranted,” said Intel’s general counsel, A. Douglas Melamed.

The case, Mr. Melamed said, “could have, and should have, been settled.” The behavioral changes that the F.T.C. is seeking, he added, amount to “new rules for micromanaging business conduct” and a potentially “dangerous turn for U.S. law.”

As a remedy, the commission is seeking an order that would prevent Intel from using threats, bundled prices, or other offers to encourage exclusive deals, hamper competition, or unfairly manipulate the prices of its microprocessors or graphics chips. The F.T.C. is also proposing requirements that Intel submit some business decisions for prior agency approval and agree to continuing commission monitoring of many of its practices.

Mr. Melamed said that the commission’s recommended remedial steps would unfairly constrain Intel’s pricing and marketing, hinder product design and innovation, and force the company to give away its intellectual property.

The complaint is an administrative action, which will be heard before a single administrative law judge within the F.T.C., with the trial starting next September. That trial is expected to be completed a few months later, much faster than if an antitrust case were tried in federal court, Mr. Feinstein said.

Still, the ruling by the administrative judge can be appealed to the five commissioners, who act as judges, and later to a federal appeals court — unless Intel reaches a settlement somewhere along the way.

The F.T.C. staff filing is a 24-page enumeration of Intel’s reported anticompetitive acts. It contains no quotes from seemingly incriminating e-mail messages or notes of conversations between Intel executives and personal computer makers, as did the New York State complaint, for example.

But the F.T.C. accusations do extend beyond the charges in other pending complaints, which have focused mainly on claims that Intel has systematically used large rebates and co-marketing arrangements to persuade computer makers to use its microprocessor chips instead of those made by A.M.D.

Besides those charges, the F.T.C. complaint accuses Intel of taking a series of steps to hinder competition in the market for graphics processing chips, which are increasingly important in running video and movies on computers. Some graphics chip makers, including Nvidia, are even building heavy-duty corporate computers from such chips.

Has the F.T.C. Opened the Door for the Great Chip War?
By ASHLEE VANCE

As part of its lawsuit against Intel filed on Wednesday, the Federal Trade Commission proposed a wish list of things it would like to see happen if it can prove its case. And one of these items is a real doozy.

I present to you relief measure No. 17:

Requiring Intel to make available technology (including whatever is necessary to interoperate with Intel’s CPUs or chipsets) to others, via licensing or other means, upon such terms and conditions as the Commission may order, including but not limited to extensions of terms of current licenses.

For those of you keeping score at home, this remedy could pave the way for a chip war the likes of which we’ve yet to encounter.

Intel and Advanced Micro Devices have set out on a path to combine computing and graphics functions on a single piece of silicon, instead of selling two different chips to handle these functions. That’s nice for Intel and A.M.D. because they both have main (computing) chip and graphics chip expertise.

Nvidia has been looking like the odd man out in this scenario, since it only makes graphics chips. As it stands today, Nvidia does not have a license to make the x86 computing chips it would need to compete against Intel and A.M.D. in this new era of hybrid products.

Parse through the legalese in relief measure No. 17 and you find the F.T.C. opening the door for Nvidia to obtain an x86 license from Intel. That’s big news, since Intel had very little motivation in the past to grant Nvidia such a license. With an x86 license in hand, Nvidia could go the hybrid route as well, and the world would end up with not two but three chip powerhouses going at it for mainstream computing devices.

Relief measure No. 18 puts an even finer point on the matter. It reads:

Prohibiting Intel from including or enforcing terms in its x86 licensing agreements that restrict the ability of licensees to change ownership, to obtain investments or financing, to outsource production of x86 microprocessors, or to otherwise partner with third parties to expand output.

And so, any company with an x86 license could do what it takes to produce computing chips with or without Intel’s blessing.

Making an x86 chip from scratch would be no easy feat for Nvidia. It usually takes about four years and close to $1 billion to get to the first run of such a product.

My contacts in the chip industry say it’s an open secret that Nvidia has been working on just such a chip since 2007. Nvidia has refused to comment on this type of speculation, and I’ve yet to hear any confirmation directly from an Nvidia employee. Usually when I ask about an x86 product, the Nvidia worker bees giggle and change the subject.

In June, I asked Nvidia’s chief executive, Jen-Hsun Huang, if Nvidia had an x86 chip lurking somewhere in its labs.

“If we have such a product, I haven’t found it yet,” Mr. Huang said. “It’s just not important right now.”

Mr. Huang might want to go talk to all the ex-Transmeta employees Nvidia has hired over the past couple of years.

A quick search on the professional networking site LinkedIn shows at least 70 former Transmeta employees working at Nvidia. Transmeta, of course, made x86 chips that competed against Intel in the laptop market.

Chip experts are always nice to have around if you’re a chip company. Their skills could be applied to any number of projects at Nvidia, including the company’s new ARM-based Tegra products for mobile devices.

But there’s a large group of Transmeta alumni that have rather niche skills. They’re what are known as x86 test and verification engineers.

These folks do the low-level heavy lifting to certify that an x86 chip performs as expected and really is an x86 product that can run all of the associated software with ease.

“The test is basically ‘does your chip behave exactly like an Intel or A.M.D. one’ — so it can run existing software (especially Windows) without breaking,” said David Kanter, the editor of Real World Technologies, a Web site covering the chip business. “A good x86 validation and test engineer needs to fundamentally understand all the bizarre nooks and crannies of the architecture, and be able to characterize implementations from Intel and A.M.D.”

In short, there’s basically one reason Nvidia would want these types of people around, and it’s to test and certify an x86 chip.

“The expertise to validate and test an x86 is incredibly specific and only useful to a few companies,” Mr. Kanter said. “Some of that expertise can transfer to other areas, but it’s a lot like having a Formula 1 racer driving a taxi cab.”

Nvidia’s research and development spending has risen significantly in recent years. In 2006, Nvidia spent $352 million. That figure jumped to $855.9 million in 2008. The bulk of the increase came from engineers’ salaries.

So, either Nvidia has a lot of cabbies around for fun or it’s paying these top-flight engineers to do the work in which they’ve long specialized.

If the F.T.C. gets its way, we may just find out the answer to the question.
akhir