A Worsening U.S. Condition: Memory Loss
The tale begins hopefully. A well-known figure in the electronics industry sees opportunity in a crying need of U.S. computer companies and responds by trying to start a company to fill that need.
But it ends ominously. Pierre Lamond, a partner in the venture capital firm Sequoia Capital and one of the founders of National Semiconductor Co., set out in May to gather backing for a company that would manufacture in the United States the semiconductors called memory chips.
Memory chips are essential for computers, but U.S. production dwindled in recent years as American companies found it unprofitable to make them. The result has been a shortage of vital parts affecting every U.S. computer company--except IBM, which makes its own chips.
Lamond saw opportunity. He could raise $30 million in venture capital and lease $125 million worth of equipment, if he could get a commitment from computer makers to buy his eventual production. He approached Digital Equipment, Apple Computer and Sun Microsystems for backing. But even though all three were short of essential memory chips, they wouldn’t back Lamond’s venture.
Why not? Lamond is blunt: “They were afraid of the reaction from their Japanese suppliers if they supported a U.S. effort to make memory chips,” he says. They were afraid their supplies would be reduced or cut off.
None of the computer companies will comment directly on the Lamond venture. All say they support U.S. manufacture of memory chips and express varying degrees of concern. “We don’t like to see any single country or company owning too dominant a position,” says a spokesman for Digital.
“Our suppliers are diversified,” says a spokeswoman for Sun Microsystems, whose business is constrained by its inability to get sufficient supplies of memory chips. “Unfortunately all are Japanese suppliers.”
Dependence Isn’t Smart
Why unfortunate? Isn’t it a global economy? What’s wrong with depending on Japanese suppliers?
The fact that in doing so U.S. computer makers are depending on their direct competitors--and that’s not smart. Toshiba, Hitachi, Fujitsu and NEC--Japan’s leading memory chip makers--are also its big computer makers. Toshiba’s laptop computer already leads the U.S. market. U.S. companies are aware that their supplier-competitors could limit access to the latest and best semiconductor parts, yet they go on buying.
Also, U.S. dependance is unwise because the American customer no longer calls the tune. At roughly $19 billion a year, Japan is now the biggest market for semiconductors, not the United States--at $14 billion. Furthermore, demand is growing faster in Japan and will continue to do so as Japanese industry pioneers such advances as high-definition television, says Michael Boss of Dataquest, the electronics industry research firm. “Should U.S. computer makers feel paranoid?” asks Boss rhetorically. “Maybe not, but they should feel concerned.”
How did U.S. companies get into this bind? By thinking only of the cards before them, like poker players, rather than two or three moves ahead like chess players.
The memory chip is a paradoxical product. It is the most numerous of the various electronic chips in most modern products, from talking dolls to cars and computers.
But precisely because they’re produced in volume, memory chips became commodities--low-priced, low-profit and ultimately no-profit products. Japanese companies flooded the market with below-cost memory chips in the early to mid-1980s, causing U.S. producers to lose $2 billion and persuading many to stop making memory chips.
Meanwhile, Japanese companies sustained collective losses of $4 billion but continued to make memory chips. Why? Because with most of their U.S. competitors knocked out, Japanese companies could get 85% of a world memory chip market that soon will be worth $10 billion in annual sales.
Also they persevered because memory is the key to electronics manufacturing. Learning how to turn chemicals and gases into millions of tiny chips of densely packed electrical circuits gives companies the ability to discover other things--including the processes that could guide the computers of the future.
Thus the humble memory chip, far from a profitless waste, is a key to technological leadership--as Makoto Kuroda, Japan’s former Minister of International Trade and Industry, understood when he declared recently: “The semiconductor is not a commodity. It is the business of the future.”
The story is not quite over yet. The U.S. could maintain a technological presence because Texas Instruments in Dallas, and Micron Technology, of Boise, Ida., still produce memory chips. And Motorola, after dropping memory production two years ago, is now getting back in with a new plant in Phoenix.
But getting back is neither easy nor cheap, and the fact that Pierre Lamond couldn’t attract backing for his company may indicate that U.S. industry is folding its cards in yet another game.
