Computer chip chase with Japan has US mobilizing for 1988 win
Washington — American companies are on the verge of launching an unprecedented joint-research program to leapfrog the best Japanese technology in computer chips.
The goal: a 4 million-bit computer chip that could seize the lead for US firms by 1988.
The research effort, which could involve more than 20 corporations, is being strongly encouraged by the US Commerce Department. It may also eventually get financial help from the Defense Department. If successful, it could:
* Help the United States regain its traditional lead in the vitally important computer-chip industry and slow or reverse the loss of business to Japan.
* Continue reducing the price of computers and other memory-related devices that are now being used in more and more consumer products such as autos, televisions, watches, calculators, video games, and even popcorn poppers.
* Ensure US military planners access to the most advanced equipment for use in aircraft and other weapons systems in the 1990s.
Spearheading the 4-megabit chip program would be the Semiconductor Research Corporation. SRC is a growing industry group that includes some of the real ''blue chips'' of the electronics industry - companies like IBM, Intel Corporation, General Electric, and Advanced Micro Devices Inc.
Final details of the project are still being hammered out. But if it goes as expected, industry insiders say, the program would bypass the next two generations of computer-chip technology with the goal of giving the United States a two-year lead over Japan.
All of this can be critically important to the future of the American economy.
Computer chips are at the heart of some of the most advanced (and profitable) products in the world today. They are vital to computers, robotics, aerospace technology, communications systems, and ''smart'' weapons.
The US has traditionally been the leader in computer-chip production and technology. But several years ago, Tokyo singled out the industry as an area where Japan could become the dominant force.
The Semiconductor Industry Association, an American organization, estimates that from 1976 to 1982, Japanese semiconductor companies received government aid totaling between $507 million and $2 billion.
The results have been devastating to US companies.
In the early days of computer-chip manufacturing, a large number of American producers were in the business. Today, few dare to compete with the Japanese juggernaut.
The first generation of computer chips, the 1K dynamic RAM, which came on the market in about 1970, was capable of storing some 1,000 bits of information. (RAM stands for random-access memory, in which the stored information can be changed by the user.) Fourteen US firms produced this size chip. Eight Japanese companies did.
In 1973, the next generation, the 4K chip (some 4,000 bits of information), was produced by 15 American firms and only six Japanese companies. The 16K chip, a more complex device that became widely available in 1978, was produced by 12 US companies, six Japanese.
Then, in 1981, Japanese companies jumped full force into the next generation, the 64K chip, which today is the most important computer-chip product on the market. Because of severe Japanese price cutting (helped by government subsidies) and excellent quality controls by the Japanese firms, only five US firms got into 64K production, compared with six Japanese companies.
It was estimated that while US firms controlled about two-thirds of the 16K market, Japan grabbed about 70 percent of the 64K market.
Japan also appears to be ahead in marketing the next generation, the 256K chip, which is just coming into production. It is to reverse this American decline that SRC, Commerce, Defense, and other interested parties have acted.
Producing the 4-megabit chip on a rapid timetable will be a challenge - both technical and financial.
Technically, the challenge is to squeeze millions of electronic connections onto a fingernail-size silicon chip. This will require creating connections less than one micron in width (a micron is 1/100th the width of a human hair). Some insiders say that creating a 4-megabit chip could require getting down to widths no greater than one-half micron.
Current chip-producing equipment, made by such firms as Perkin-Elmer, GCA Corporation, General Signal Corporation, and Japan's Nikon, cannot produce connections that fine on a day-in, day-out production basis. Even the best, under factory conditions, are widths of only 1.2 to 1.5 microns.
Under laboratory conditions, however, some of today's equipment has produced connections with widths as narrow as 0.5 to 0.8 microns. The challenge will be to do that on a production line.
Insiders say that may require one of several developments: improvements in today's best technology, which uses ultraviolet light to draw the fine lines on silicon; new methods that combine today's technology with X-rays; or going beyond today's methods to holography or electron-beam equipment.
The choice of methods will be pivotal. If SRC chooses a version of today's technology, it will reduce the risk of the project. The 4-megabit chip will probably be produced in time. The US could regain its worldwide lead.
Pushing today's technology to the limits to produce the 4-megabit chip, however, could be self-defeating. It may be better, insiders say, to increase the risk by trying to create new production equipment that could not only produce the 4-megabit chip, but also even more advanced chips in later years.
There is also the question of financing. The cost will be near $200 million, it is estimated. SRC hopes $50 million of that will come from the Defense Department.
The balance could come either from corporations involved with SRC, or from private venture capital - ''people like doctors, lawyers, dentists'' - who could invest through special tax-shelter provisions, a Commerce official says.
If the project succeeds, it could not only benefit these investors, but also put some cash back into the pockets of US chipmakers.