Smaller Pond, Bigger Voice : Fiber-optics: Sanford Kane, former IBM executive, presides over the second effort to expand sales at tiny PCO Inc.

A year ago, it looked like Sanford L. Kane might head up a $1-billion consortium of American companies to manufacture computer chips. But the consortium, U.S. Memories, never got the money it needed to get off the ground. Now, Kane, a former longtime International Business Machines executive, is at the helm of a decidedly more modest consortium: PCO, Inc., a Chatsworth-based joint venture of IBM and Corning Inc. with $15 million a year in sales and no profits to show for it.

PCO’s board of directors decided it needed someone new in the top job, and Kane’s mandate is to to turn the tiny company, which makes components for transmitting voices and data on fiber-optic cable, into a much bigger and profitable manufacturer in the growing fiber-optics business.

“Kane was the obvious choice,” said James C. McGroddy, vice president and director of research of IBM, which acquired 25% of PCO from Corning in early 1989. McGroddy was one of two PCO board members who headed up the search for a new chief executive.

But a big expansion in sales is hardly a new idea at PCO. Nearly two years ago the company started gearing up for big sales growth and nothing happened.

PCO’s overly optimistic sales projections weren’t without costs. In the last month, the company has laid off about 30 workers, some of whom were temporary employees, reducing its payroll to around 150. And PCO has been losing money, said Larry Aiello, a member of the company’s board and director of planning and development with Corning, which owns 65% of PCO.

Kane said the lack of growth wasn’t PCO’s fault; the widely expected surge in demand for PCO’s equipment simply never materialized. Now Kane said PCO again “is counting on a lot of increases” in sales of its components--especially to IBM--but this time, Kane said, things will be different.

Fiber-optics uses glass threads as thin as human hairs to carry signals by means of light pulses. A cable made of optical fibers can carry hundreds of times more information than the same thickness of normal copper wiring--and over longer distances without as much help from devices that boost weak signals.

PCO makes “optoelectronic interfaces” that act as transmitters and receivers of the light pulses, converting them into electronic signals used by computers and phones.

Long distance phone companies like AT&T;, MCI and Sprint have already wired the country with fiber-optics, vastly increasing the capacity of their telephone systems. But fiber-optic systems can also carry computer data, enabling peripheral devices, like terminals, to be placed greater distances from the main computer. Fiber-optics can even reduce the clutter from coils of copper wiring within computers. It’s this kind of data communications that PCO is looking to for its growth.

Without its parents to back it up, PCO would seem like the scrawniest kid on the block among fiber-optic component makers. Many of PCO’s competitors in the field are big Japanese companies like Mitsubishi, Fujitsu and Sumitomo. Siemens, the German electronics company, and AT&T; are also important players.

Two years ago industry observers were predicting a revolution in the use of fiber-optics for carrying computer data. But in fact, according to ElectroniCast, a San Mateo market research firm, 1990 U.S. sales of devices like the ones PCO makes will probably be unchanged from a year earlier at $320 million. ElectroniCast’s Mike Osterman said that the number of such devices sold increased about 23%, but that the increase was offset by steadily falling prices.

One sign that the market might finally be ready to grow came recently from IBM. In announcing the introduction of its newest line of big mainframe computers early in September, the company said that it would give buyers the option of connecting the mainframes to peripheral devices via fiber-optic lines.

And IBM could play a big part in making PCO’s current sales projections come true. In fact, Kane said his IBM ties, developed over 27 years at the company, are “going to be critical to the future as IBM becomes a large customer of PCO’s.” In fact Kane said one of his jobs is to make sure PCO meets IBM’s standards for its suppliers.

As a big customer as well as a 25% owner of PCO, IBM could be very influential at the tiny Chatsworth company. Indeed, Ulrich Weil, a Washington-based financial analyst, said, “Traditionally, when IBM goes into something, they sooner or later put their man in the top.”

But Aiello denied any suggestion that “IBM forced Sandy Kane on PCO.”

PCO was founded in 1984 by Michael Barnoski, a former research executive with TRW and Hughes Research, along with Corning Inc. and Plessey Co. of Britain. In 1986, Plessey sold its interest to Corning. And with the company’s sales growing steadily from scratch to $15 million in 1989, IBM bought its 25% stake for an amount reported to be $10 million.

But this year Barnoski agreed that PCO needed a new leader, according to McGroddy, and with Kane’s appointment, Barnoski will become vice chairman of the company’s board, heading its long-term research and development.

Taking over at PCO is Kane’s second fresh start in as many years. In June, 1989, the former division vice president of industry operations at IBM became chief executive of U.S. Memories, an attempt by leading American computer makers to form a manufacturing alliance to grab back from Japanese companies a chunk of the market for dynamic random access memories, or DRAMs, which are basic memory chips for computers.

Kane’s job at IBM had made him the company’s liaison to the semiconductor industry, and he’d helped lead a domestic computer chip research and development consortium.

U.S. Memories was given a small amount of start-up money by computer makers IBM, Digital Equipment Corp., Hewlett-Packard Co., and chip makers Intel Corp., National Semiconductor and Advanced Micro Devices. The idea behind U.S. Memories was to raise $700 million to build a plant to make DRAMs (pronounced dee- rams) to fight off the Japanese sales in this country. Kane hoped to get about $150 million from computer companies.

U.S. Memories was supposed to not only give the United States a foothold in the world DRAM market, but give the computer companies that invested in it control over a portion of their chip supply. To those companies, said Kane, “I wasn’t waving the flag at all. . . . This was purely a business discussion.”

But not all computer and chip makers liked the discussion. T.J. Rodgers, president of Cypress Semiconductor, a Silicon Valley chip maker, was U.S. Memories’ chief critic, calling it “a bad idea, and not necessary.”

In the mid-1980s, Japanese manufacturers nearly wiped out the U.S. industry with a flood of low-priced DRAMs in a move some Americans complained was unfair. By 1989, however, there was a shortage of the crucial memory chips, demonstrating that U.S. companies were now at the mercy of foreign manufacturers. But Kane figured that by 1995, U.S. Memories could produce more than $1-billion worth of chips a year.

Kane’s biggest task was not finding investors, but getting computer manufacturers to guarantee that they would buy up most of U.S. Memories’ chips from its first manufacturing plant. With those guarantees, Kane thought getting other investors and lenders would be no problem. But without them, the task was futile.

By last January, however, it was clear that only IBM and Digital were willing to make such commitments, and the effort fell through.

Kane spent about six months closing down U.S. Memories, then started looking for a new job. Of three offers he received, Kane said PCO’s was the only one that made him chief executive, a position that appealed to him. Kane figures he could have gone back to IBM, too, but “I didn’t want to,” he said. “As much as I liked IBM, I wasn’t going to run the place.”