January 18, 1990
"High-tech Firms Saw Uncertainty in U.S. Memories"
San Jose Mercury News
By Timothy Taylor
<< Back to 1990 menu

JUDGING from the post mortems, the official cause of death for U.S. Memories, the semiconductor start-up that never quite started up, was a failure of computer firms to cooperate.

But while companies certainly did not flock to invest their money in U.S. Memories, it's worth remembering that what looks like uncooperative behavior to one person may look like a sound business decision to someone else. After all, as semiconductor industry executives have been eager to point out when appealing for government assistance, the U.S. chip business has been limping badly in the 1980s. And cooperation is not a cure-all; joint ventures have no special guarantee of success.

The original concept of U.S. Memories was that chip-making and chip-using companies would raise $1 billion (!) to start a new semiconductor company, which would guarantee an inexpensive and plentiful supply of high-quality, technologically advanced chips to its member computer companies. Those companies would put up money and guarantee to purchase about a third of their chips from U.S. Memories. The seven original members of U.S. Memories who announced this proposal last June were Intel, National Semiconductor, Advanced Micro Devices, LSI Logic, IBM, Digital Equipment Corp. and Hewlett-Packard.

The idea sounded pretty good in the press releases last summer, but the high-tech industry has seen a lot of promising business plans wither and die. In this case, companies like Apple, Sun Microsystems, Unisys, Tandem and NCR all eventually decided against investing in U.S. Memories, and they are now being blamed for their decision.

But these companies didn't decide against cooperating with the U.S. Memories scheme out of sheer cussedness. They too have had to live with semiconductor prices and availabilities swinging wildly from boom to bust. They are aware of the dangers of becoming too dependent on a few foreign suppliers of chips. I'm sure they would love to see several more healthy U.S. semiconductor companies jousting to supply them with chips.

However, they also had to weigh the possibility that U.S. Memories might not be the answer to their prayers. What if the venture didn't succeed in producing high quality at competitive cost? Remember, U.S. Memories was also requesting that they guarantee to purchase its chips. A new generation of semiconductors seems to come along every two years; what if U.S. Memories fell behind in future generations?

There was surely some reason to question whether U.S. Memories would be successful. After all, executives in the U.S. semiconductor industry have spent a lot of time recent years explaining at high volume that U.S. chip makers have a number of built-in disadvantages in competing with the Japanese.

One recent report describing many of these disadvantages is "A Strategic Industry at Risk," from the National Advisory Committee on Semiconductors. The committee was set up by Congress but its membership is heavily weighted toward executives from the chip industry.

One problem it cites, for example, is that the lucrative consumer electronics market has been taken over by companies in the Far East. "Looking toward the year 2000, Japanese sales of semiconductors for consumer products are projected to be $13.5 billion greater than those of U.S. chip producers. If this projection is accurate, consumer products could fund more than $1 billion of R&D in Japan in excess of that being funded in the United States."

A second problem is the high cost of capital in the United States. This problem hurts all U.S. manufacturing industries to some extent, but with the cost of a state-of-the-art semiconductor plant now in the hundreds of millions of dollars, the chip business is hit especially hard.

Again, the Committee on Semiconductors: "While there is some debate over the precise values for the real cost of capital in the United States and Japan, there is agreement that costs in Japan have been significantly lower. This differential was particularly important during the early part of the 1980s when the Japanese semiconductor industry was investing heavily to expand capacity." One main reason that the U.S. cost of capital rose in the early 1980s were the enormous federal budget deficits.

The committee lists many other disadvantages for the U.S. semiconductor industry, ranging from the educational system to laws affecting antitrust, trade, taxes and much else. Although I don't buy into its entire agenda, the bottom line is clear: U.S. chip makers held over 60 percent of the world market in 1980, but less than 40 percent today.

U.S. Memories sounded so attractive because it promised that with just a little cooperation, the American semiconductor industry would be restored in one fell swoop. If only restoring U.S. competitiveness were so simple!

I had hoped that U.S. Memories would succeed. But I'm not especially surprised that many computer companies chose to suffer the uncertainties of the semiconductor market from a distance, rather than commit their investment capital and future purchasing decisions to a company that never existed.

<< Back to 1990 menu