Flat-Panel Firms Struggle as U.S. Priorities Shift

When it comes to chutzpah, it’s hard to beat Harry Marshall.

As chief executive of Silicon Valley start-up Candescent Technologies Corp., Marshall has raised $150 million from the likes of Hewlett-Packard Co., Compaq Computer Corp. and the U.S. government by insisting that in the year 2000 he will have a product worth $1 billion in sales.

Here’s the catch: The product, a new form of flat-panel computer display, won’t be ready until 1998. The only prototypes built are scarcely bigger than the palm of your hand--and that’s after five years of development and an expenditure of $100 million. And the company has only begun to grapple with the myriad production problems bound to arise as it scales up to mass-produce large screens.

Many analysts doubt that Candescent will ever overcome the technical and production problems it faces. If it fails, that will all but spell the ignominious end to one of the Defense Department’s most ambitious industrial research programs.

For Candescent is about all that’s left of the Clinton administration’s four-year, $600-million effort to beat Japan at its own game: using government subsidies to play catch-up in a strategically important technology.

The rationale for the effort seemed compelling at the time of its launch. Flat-panel displays--inches-thick screens with the high resolution of computer monitors--are key components of jet aircraft and other military equipment. Japan, which controlled 95% of the market, was unwilling to supply custom components to the U.S. military. Meanwhile, a shortage of displays was hurting U.S. laptop computer makers such as Apple Computer Inc.

Kenneth Flamm, the Brookings Institution scholar who was the architect of the Defense Department’s flat-panel policy, predicted that the department would help create by the year 2000 a flourishing U.S. industry with a 15% share of a market that would then be worth $20 billion.

The program’s promoters also hoped the effort might enable American companies to leapfrog their Japanese competitors with cheaper and better displays. The standard technology used in the field--known as active matrix liquid crystal displays, or AMLCD--was then, and remains today, complex and expensive.

Each screen contains hundreds of thousands of transistors that “switch” liquid crystals sandwiched between two plates of glass. When the crystals are aligned, light passes through a filter creating a colored dot on the screen. Production is delicate: One defective transistor and the whole panel must be discarded.

Consequently, the Defense Department financed the newly established U.S. Display Consortium, based in San Jose, by ladling out millions in government grants.

AT&T;, Xerox and a Wisconsin-based LCD manufacturer named Standish Industries won a combined $50 million in matching grants to build a pilot plant to manufacture AMLCDs using new Xerox technology.

Candescent, then called Silicon Video, won $40 million to develop its own technology, which the company claimed would produce panels brighter and clearer than AMLCD at 30% less cost.

But, as the grant recipients were starting their work, the policy was already foundering in Washington.

One year after the program’s launch, the new Republican majority in Congress forced the Defense Department to cut back its annual spending on flat panels to $50 million.

Meanwhile, Japanese and South Korean companies pulled far ahead of their American counterparts by pouring billions of dollars into new factories packed with efficient new equipment.

Panel prices plummeted, eliminating one major concern of the U.S. planners.

“The whole premise of government policy was based on AMLCD being too expensive,” says Lawrence Tannas, an LCD expert based in Orange. But with hundreds of companies working on improving the technology, says Tannas, incremental improvements could keep it ahead of any new developments.

South Korea’s entry into the industry also alleviated the Defense Department’s fears of a shortage of willing suppliers.

With displays cheap and readily available, U.S. multinationals’ interest in the technology rapidly cooled.

Xerox’s Palo Alto Research Center--the legendary Xerox PARC--had developed a way to make LCD panels as clear as print on paper. It spun the technology off into a new subsidiary called dpiX with a focus on serving small high-end markets.

Motorola, which invested large sums in company in the Portland, Ore., area called Motif in hopes of getting a supply of panels for its portable communications products, pulled the plug after the start-up ran into production problems.

“Government can help build up a company’s technological capabilities,” says Jon Clemens, a senior research executive at Sharp Corp., the world’s flat-panel display leader with 40% of the market, “but making companies more competitive isn’t something the government does very well.”

Perhaps the leading case in point was AT&T.;

AT&T; put together a talented team of 20 top researchers and signed a joint production agreement with a Japanese supplier. To develop production experience, AT&T; built a state-of-the-art production line funded in part by $15 million of Defense Department funds.

But a year later, after a group of AT&T; engineers had already been sent to Japan for training, AT&T; executives began to waffle. Their videophone products were doing poorly in the marketplace, and the company was abandoning its foray into the personal computer business. That removed two potential internal customers for the panels.

AT&T; withdrew from the Japanese deal. As late as April 1995, AT&T; executives were still asking themselves in a technical meeting whether it made sense for the company to build flat panels since AT&T; was “never good at high-volume manufacturing,” an insider recalls. When a researcher asked why AT&T; couldn’t address its production problems, the insider says, the reply came from Arno Penzias, a Nobel Prize-winning senior executive at Bell Laboratories (now chief scientist at AT&T; spinoff Lucent Technologies).

“This is AT&T;,” Penzias reportedly said. “At AT&T; we don’t believe in ourselves.”

A Lucent spokesman said Penzias was unavailable to confirm that account of the conversation.

Early this year, soon after announcing it would split into three companies, AT&T; departed the flat-panel business, donating most of its equipment to two universities and firing the 20 researchers working on the project.

“It’s pathetic,” says Peter Brody, who invented AMLCD while at Westinghouse in the 1960s only to watch as company after company declined to commit to manufacturing the product. Since most telephones of the future will include an LCD screen, says Brody, “AT&T; should have invested in this technology. They chickened out.”

Today, to survive, American flat-panel manufacturers are targeting niche markets in medical and military fields.

But those don’t demand the high-volume manufacturing needed to support the infrastructure of equipment and materials suppliers the Defense Department hoped to create.

What’s left are companies like Candescent, which hopes to find the $400 million it needs to move to mass production by partnering with two major manufacturers. If the past is any indication, that won’t be a cakewalk.