US Chipmakers Try to Win on Uneven Field

CALL it high technology's new realism. After a long and unsuccessful battle to make world trading rules more fair, United States chipmakers are switching strategies. They are positioning themselves to compete better in the current world market instead of trying to change it, a new survey concludes.

"There's the recognition that the playing field isn't level and isn't going to be level in the near term," says Steven Burrill, a technology analyst with Ernst & Young, which released the survey Friday. "What the survey shows is how effective we are becoming as an industry in learning how to compete in a nonlevel playing field."

"Knowing your competitor is the first step," agrees Gerardo Ungson, director of the international business program at the University of Oregon. His research suggests US companies have become much more aware of Japanese business practices in the last few years.

There are hints this knowledge is paying off. For the first time since 1979, the US semiconductor industry gained world market share last year. A few companies are maintaining, or even increasing, their technological lead.

In the next five years, they aim to stress innovation even more, the survey found. Mid-size to top-tier companies plan to maintain or increase slightly the share of revenues they invest in research and development. (On average, the industry spends 21 percent of revenues on R&D compared with 8 percent for electronics companies in general.) Nearly 70 percent of the semiconductor firms said they are shifting their R&D money from basic to applied research.

The Ernst & Young report surveyed top officials at about one quarter of US semiconductor companies. Producing a wide array of products from memory and logic chips to microprocessors that lie at the heart of computers, the industry is considered crucial to the nation's high-tech competitiveness.

Besides stressing innovation, semiconductor companies aim to improve their manufacturing and marketing flexibility and to focus on the areas of the market where they're strongest.

That's good, as far as it goes, Professor Ungson says. But to compete in the existing world market, chipmakers will have to work in concert as an industry. So far, there's much disagreement - and even less coordination - on what strategies the US should pursue. For example, the industry can't make up its mind about Sematech, a consortium of US semiconductor makers who receive federal funding to improve chip manufacturing.

Andrew Grove, president of the successful Intel Corporation, praises it. "We have to continue to pull together where and when we can," he said in a speech last fall.

Others criticize it as a government boondoggle. Many others doubt it will work. "I think it would be a terrific thing if we could pull it off, but I am very skeptical," says Michael Hackworth, president and chief executive office of Cirrus Logic. "The history has not been that the US people know how to do that."

INSTEAD of the extremes of the consortia and the go-it-alone approach, US semiconductor companies are taking a middle road and forming strategic alliances, Mr. Burrill says.

It is widely acknowledged that the first wave of these alliances, often with foreign firms, didn't work. Often, US companies gave away valuable technology for short-term infusions of capital. Some chief executives, such as Robert Swanson of Linear Technology, argue that these alliances still don't work.

But Burrill and other executives argue that US companies have learned from the past and are making new alliances that fairly link US innovation with other countries' manufacturing expertise. Already, 80 percent of the companies surveyed were involved in some type of manufacturing agreement; 85 percent expect to have such an agreement by 1995.

These alliances are small entities compared with the industry consortia in Japan and Europe and the cross-industry groupings such as Japan's keiretsu and South Korea's chaebol, which provide powerful support to companies.

US trade policy is another area where the US semiconductor companies could act in better concert.

There's wide agreement about the problem. "We can't live in this nostalgic past that the greatest principle to live by is free trade," says Robert Swanson, president and chief executive officer of Linear Technology Corporation. "The rest of the world is not playing like that."

But there's a wide divergence of opinion about what to do about it.

Several large US chipmakers, such as Advanced Micro Devices, argue that the soon-to-expire semiconductor agreement between US and Japan is the only pact that has come close to working. But the Ernst & Young report barely touches on it. Some chief executives suggest that, despite a dramatic loss of world market share during the 1980s, the current US model will work in the 1990s.

The US-Japan semiconductor agreement is significant because it moves in the other direction. Instead of pushing for a freer market, the 1986 pact promised that non-Japanese firms would get 20 percent of Japan's semiconductor market. Although they have gained market share in Japan, the US says that share is still below 20 percent.

According to published reports, the US and Japan agree on the general outlines for a new plan to replace the current agreement, which expires in July. But they disagree on important details, and did not reach a settlement in the latest round of talks in Tokyo Friday and Saturday.

If the US really wants to compete in this new world market, its government will also have to make institutional changes, Ungson says.

Industry executives regularly call for a lower budget deficit, tax policies that favor investment and research, and improvements in education.

Significantly, 56 percent of those surveyed pointed to the lack of government support as a US weakness. They mentioned it four times more often than any other factor.