That was the question many asked when Intel reported in its third-quarter earnings Tuesday that it would pay a $300 million charge to settle a lawsuit with MicroUnity, an eight-employee company in Santa Clara, Calif.
The deal, which brought an end to a patent-infringement suit filed by MicroUnity in March 2004, was the first some had heard of the company, which specializes in media processors. But for Silicon Valley veterans, the MicroUnity name conjured up memories of the high-tech business before the Internet era, when Packard Bell and Egghead Software strode the Earth and Google founder Larry Page was just another guy in high school.
"They were the Transmeta of their day," Insight 64 analyst Nathan Brookwood said. Chipmaker Transmeta came on like gangbusters several years ago, promoting an energy-efficient processor for laptops, but ran into manufacturing problems and the dot-com bust.
In its suit, MicroUnity alleged that Intel had infringed on its patents with its use of extensions for multimedia and application threading, a process that lets a chip perform two tasks at once.
The settlement underlines the trend for small companies to seek revenue through patent lawsuits. In recent years, for example, Intel spent $675 million to put an end to a similar action by Intergraph. In MicroUnity's case, the deal illustrates how legal action can revive the promise of a once-hot company.
Founded in 1988 and drifting toward obscurity by 1997, MicroUnity was one of a number of companies touted as the next Intel. Its strategy revolved around developing chips that could process audio, video and similar streams of data inside TVs and other consumer electronics. Some theorized at the time that success in these new markets could shift the balance of power away from the PC.
"TVs, telephones and radios were all going to become digital," MicroUnity founder and CEO John Moussouris told CNET News.com on Wednesday. "We asked, 'What should a computer look like that can do the real work of a TV?'"
While other start-ups pursued the same strategy, MicroUnity had a better pedigree than most. A graduate of Harvard University and a Rhodes Scholar, Moussouris was also one of the founders of MIPS Computer Systems, which made the MIPS chip used inside servers and workstations from Silicon Graphics Inc.
A decade ago, SGI's machines were celebrated for their performance, and the company landed many of the major contracts in Hollywood. (Other MIPS alumni from the era include Stanford president John Hennessy and Forest Baskett, a partner at New Enterprise Associates.)
Millions in venture money flowed in from Microsoft, Hewlett-Packard, TCI (Now AT&T Broadband), Time Warner, Cox Communications, Motorola and Comcast, among others. Moussouris' Harvard roommate, Will Hearst, now a partner at Kleiner, Perkins, Caufield & Byers, was an early investor.
Moussouris published articles in several prominent publications. MicroUnity also opened its own chip fabrication facility, a luxury most processor companies can no longer afford.
Media processors, however, never fully caught on. General-purpose PC chips steadily became more powerful, thanks to the increasing number of transistors they packed, in keeping with Moore's Law. The dominant position of chips based on x86 architecture, like the Pentium and Opteron families, also became solidified, as computer manufacturers and software developers showed less and less interest in adapting their products to novel chip designs.
Media processors weren't cheap to design or produce either.
"When you were looking at a $300 to $400 media processor to power a cable modem, suddenly it wasn't making as much sense," Brookwood said.
With the PC on its way to becoming a fixture in the digital home, opportunities began to dry up for chip companies like MicroUnity. Some, such as ArtX, were sold; others crumbled. MicroUnity sold a unit to lithography giant ASML and licensed chip technology to others. However, it only made prototypes of its processors and never sold them commercially.
Linley Gwennap, president of the consulting firm Linley Group, added that MicroUnity also seemed to suffer from an excess of ambition.
"If you want to talk about innovative, they were off the charts. They were looking at making major changes to the microarchitecture and fabrication techniques. They just had too many radical ideas in one package," Gwennap said. "But from a technology standpoint, they had some really innovative stuff."
Moussouris to some extent agrees, saying that the company sometimes has suffered from an "attention excess disorder." Still, it was necessary, he believes.
"The original vision of turning TVs and radios into agile computers that can change their behavior when software is downloaded--that is a challenging problem. It takes decades of effort," he said.
The $300 million settlement will likely allow MicroUnity's investors to see a return. Although sources say the company was one of the biggest recipients of venture capital in the old days, the total investment was less than half the settlement amount. Moussouris declined to comment on any future plans or offer details of how MicroUnity will use the settlement.
Other lawsuits, however, could follow. The Intel suit involved eight patents, but MicroUnity has 66 patents and patent applications, according to the U.S. Patent and Trademark Office. Those could become the basis of legal actions against other companies.
The Intel settlement effectively insulates the chip giant against further lawsuits, and it protects PC makers using Intel chips. Under the agreement, Intel will pay $140 million to MicroUnity now and the remaining $160 million over 10 years.