TOKYOThe Semiconductor Industry Association is raising questions over the Japanese government's multimillion-dollar plan to fund research aimed at busting through fundamental chip technology barriers and reviving Japan's semiconductor industry. The moves could reopen some old wounds in U.S.-Japanese semiconductor trade relations.
Although it has not taken a formal stance on the issue, the SIA has met with U.S. government officials overseeing U.S.-Japanese trade over the research project and is keeping a careful eye out for potential violations of open trade rules.
"We've been aware for some time that there is a proposal for such a project, but we've not formed any opinions on it, because we don't know a whole lot about the project," said Roger Mathus, director of the SIA's Japan office.
The SIA's chief worry is that the project will cover product-specific technologies while excluding U.S. companies from participating. Such a scenario conjures up the specter of Japan's VLSI project of the 1970s, which many on the U.S. side believe helped vault Japanese companies into DRAMs and precipitated the U.S. chip industry's downfall in the 1980s, leading to a trade war.
Japanese industry observers, however, say the concerns are overblown, pointing out that direct government funding in semiconductor R&D projects is considered normal in the United States and that Japan should not be held to a double standard.
At issue is a Ministry of International Trade and Industry (MITI) plan, approved in December, to spend about $160 million on a super-clean-room project that will construct a new R&D facility to develop semiconductor technology at 0.1 micron and below. Takeshi Kawamoto, deputy director of MITI's machinery and information industries bureau, said the only thing that's been decided is that a clean room will be built in Tsukuba.
"Japan has not had a national laboratory for semiconductor development. Isn't this fact rather strange?" he said.
Funding for the project passed the Diet in December as part of a special supplementary budget. The clean room will begin operation in fiscal 2002 at the earliest.
The construction work has already been passed to the Ministry of Construction. Blueprints are expected to be ready by October. Other details about the R&D lab are still being worked out. Kawamoto said it has not yet been decided which companies will be involved, what will be studied or how the R&D center will be managed.
MITI now funds a number of national laboratories through the Agency of Industrial Science and Technology. Fifteen labs pursue such subjects as measurement, material, machinery and electronics. The labs will be reorganized under one independent legal entity by April 2001. The proposed clean room will be positioned as the core of the new organization, with its main purpose being the pursuit of next-generation semiconductor technology.
The clean room is seen as a launch pad for closer cooperation among industry, academia and government on a wide range of R&D projects using supplementary budget funding.
A closed club?
According to one knowledgeable source, the SIA has expressed concern to the U.S. government that U.S. chip makers will be excluded from the project, which in its view would be tantamount to a direct subsidy for the Japanese semiconductor industry aimed at undercutting its U.S. counterpart. "If someone is trying to raise their own boat and torpedo another boat, that can be a problem," the source said.
The U.S. government has not taken a position yet and has not raised the issue with MITI. The United States intends to wait until the SIA polls its members before deciding how to respond, the source said.
SIA president George Scalise declined to comment on the situation. But group chairman T.J. Rodgers of Cypress Semiconductor said no one has officially raised the issue at any SIA meetings. "Japan's [government] technology programs are why their economy is screwed up now," he added, claiming that the United States is not unduly perturbed by the MITI project. "MITI's programs have been consistent catastrophies."
Even so, the U.S. government and SIA will review World Trade Organization rules to see whether Japan has violated the trade pact by giving preferential treatment to domestic companies. Proponents of the Japanese plan, however, point out that the funding targets pre-competitive research and development, not specific products, and therefore should not violate WTO rules.
The United States and Japan only recently closed the book on their long-running spat over foreign chip makers' market share in Japan. The dispute resulted in the formation of the U.S.-Japan Trade Agreement in 1988. After years of steady increases in market share by U.S. companies and the establishment of the multinational World Semiconductor Council, the trade agreement was dissolved last year.
One possible way to avoid a new fight would be to open the semiconductor research project to foreign and domestic companies. Some U.S. companies, including Motorola, Texas Instruments and IBM, already blur the line between foreign and domestic because they have substantial sales, engineering and manufacturing operations in Japan.
"The cost of development is getting higher and higher, and we have to compete with more companies outside of Japan," said Toshiaki Masuhara, senior chief engineer for Hitachi's Semiconductor and Integrated Circuits division (Tokyo). "Closing Japan doesn't make sense."
Many companies here are nonetheless troubled by the rate at which Japan's chip industry has fallen behind the United States and are betting that improved cooperation among industry, government and academia will reverse the downward spiral. Japan's worldwide chip market share now stands at half of what it was at its apex, in 1988, when it accounted for 40 percent of the global market.
"Many people believe the Japanese semiconductor industry is in crisis, so we have to do something to compete with the U.S. and other Asian countries," said Taro Okabe, executive researcher and acting executive director for the Semiconductor Industry Research Institute Japan, a chip-industry think tank.
Further, the United States should look in its own backyard before accusing Japan of favoring its domestic industry, Okabe said. The U.S. government provides funding to several national laboratories with clean rooms, such as Lawrence Livermore Labs. The government also funded Sematech-a U.S. semiconductor R&D consortium that in its early years was closed to Japanese companies-for most of that group's existence.
Others pointed out that the U.S. military has guided and funded semiconductor research through the Defense Advanced Research Projects Agency.
For many years, MITI's funding of semiconductor projects has been sporadic, but the government's attitude is changing. Prime Minister Keizo Obuchi has proposed a 120 billion yen ($1.08 billion) Millennium Project, which is separate from the super-clean-room project. The Millennium Project will focus heavily on information technology and is expected to pass the Diet in April.
Additional reporting by George Leopold and Brian Fuller.