Not in the Chips?

To the Editors:

As a research engineer in the integrated-circuit industry, I would like to comment on several factual errors in Robert B. Reich’s piece, “Japan in the Chips” [NYR, November 19].

Reich creates the impression that the Japanese have now taken the lead in IC technology. The Japanese did take the lead in manufacturing quality of one type of chip, the dynamic RAM in the late 1970s. No such lead was present in the vital microprocessor or telecommunication chips, and the Japanese are far behind in microprocessor design.

American chip makers have responded with a vigorous and successful push for quality. An independent consulting firm, Integrated Circuit Engineering, recently compared American and Japanese RAMS and found no significant quality differences (Electronics, January 13, 1982, p. 33).

Lack of demand, rather than technological inferiority, led two US firms to drop bubble memories last year. Bubbles are too slow to compete with fast, cheap silicon devices.

Reich asserts that when the recession ends, rising demand will exceed our manufacturing capacity, creating an opportunity for Japan, as occurred in 1977. Having learned their lesson, US firms have avoided the sharp cut-backs in research and capital equipment spending that the 1974 recession brought on. Capital spending for the US industry as a whole in 1981 matched that of 1980, in spite of huge drops in profits and prices during the year (Business Week, January 11, 1982, p. 77).

Reich asserts that the Defense Department’s Very High Speed Integrated Circuit (VHSIC) program will not produce technology that can be applied to commercial products. Japan’s Ministry of International Trade and Industry (MITI) has made high-speed circuits the cornerstone of its long-range plan for their computer industry. Furthermore, as Reich himself points out, development money from NASA and the Pentagon was the main reason American firms led the world in electronic technology in the 1960s and 1970s. The VHSIC research teams will be somewhat shielded from cyclical economic ups and downs, as MITI has done for Japanese computer firms.

Mr. Reich is entirely correct in pointing out that the defense buildup will aggravate the shortage of engineers, particularly those with advanced degrees. Japan graduates more electrical engineers yearly than the US, from a population half as large. The government should act to see that we are not hampered by a lack of mathematically literate high school graduates and fully trained scientists.

John J. Nangle

Dallas, Texas

Robert Reich replies:

The proof is to be found in the marketplace. The plight of America’s semiconductor manufacturers is now even more serious than it was when I wrote “Japan in the Chips” last November. Latest sales figures confirm Japanese penetration of the 64K chip, which is rapidly becoming the largest-selling memory component. In 1981, Hitachi shipped 40 percent of the 12 million 64K chips sold worldwide. The second largest market share is held by Fujitsu, which has 20 percent of worldwide sales. The largest US competitor is Motorola, with just under 20 percent. US semiconductor manufacturers together hold only 30.5 percent of the 64K chip market, compared to 69.5 percent for the Japanese. The erosion has been precipitous.

It is now clear that Hitachi will be the first electronic firm to mass-produce the next generation of computer memory chip—the so-called 256K RAM. Hitachi will begin volume sales of this high-density chip in 1983, two years earlier than most industry observers had expected. It will be producing hundreds of thousands of these chips by early 1984. This development will give Hitachi an important edge over competitors, providing a whole extra increment of sophistication in memory, and bringing substantial reductions in the price of machinery, like personal computers,that rely heavily on semiconductor components. America’s Mostek and IBM are both at least three years away from the 256K chip.

Why are we losing the chip race so quickly? For three reasons. Japanese manufacturers are producing cheaper chips. Their chips are of higher quality. And their chips embody more advanced technology. Why can’t we match this outstanding record? Because our semiconductor businesses and our government agencies that affect them (notably the Pentagon) are not organized to do so. The military buildup planned for the next few years will continue to distort this commercial market, creating bottlenecks in the production of key subcomponents and capital goods, and shortages of engineers and scientists.