Japan's semiconductor makers are also buoyed by hopes that the post-PC era will reward their expertise in powering digital consumer electronics. "A favorable wind is blowing" is a comment often heard in semiconductor executive suites.

The focus on joint R&D echoes the national VLSI R&D project of the mid to late 1970s that vaulted Japan to worldwide dominance in DRAMs. But that effort also triggered ugly semiconductor trade conflicts with the United States, causing Japan to back off large-scale national R&D projects. But now Japan's manufacturers think they have no choice but to re-create what made them successful in the first place.

Tsugio Makimoto, corporate adviser of Sony, however, warns against high expectations. "Japan actually has an accumulated technological strength. If Japan can combine its technological strength and growing demands for digital consumer electronics, it has a chance to recover. But it is not an easy task. The recovery does not come by itself. The wind may be favorable, but Japan still needs desperate efforts to recover."

Makimoto, famous for his Makimoto's wave theory, which predicted the rise and fall of certain technology generations, was the chairman of the Semiconductor in New-Century Committee (SNCC) organized in 1999. In its 2000 proposal, SNCC warned that Japan's semiconductor industry was in grave danger of collapsing and recommended that joint R&D activities involving industry, academia and government were imperative.

Ten of those companies (formerly 11 companies before Hitachi and Mitsubishi joined forces to establish Renesas), excluding Elpida, form the Semiconductors Executive Committee of the Japan Electronics and Information Technology Industries Association (JEITA) and are the investors of joint R&D companies. These include Semiconductor Leading Edge Technologies, Inc. (Selete), Semiconductor Technology Academic Research Center (STARC) and Advanced SoC Platform Corp. (ASPLA).

At the turn of the century, SNCC's proposal moved all sectors-industry, academia and government-to form a number of R&D consortia: MIRAI, ASUKA and HALCA, then ASPLA.

• Asuka: a Japanese word meaning "Advanced semiconductors through collaborative achievement."

• MIRAI: Millennium Research for Advanced Information Technology.

• HALCA: Highly Agile Line Concept Advancement.

• ASPLA: Advanced SoC Platform Corp.

These four major R&D projects operate at two bases: the Super Clean Room (SCR) Facility in Tsukuba, Ibaraki Prefecture, and the Advanced SoC R&D center in Sagamihara, Kanagawa Prefecture. Both bases were funded through a budget established by the Ministry of Economy, Trade and Industry (METI), with 25.2 billion yen (about $210 million) for SCR and 31.5 billion yen (about $ 265 million) for the Advanced SoC R&D Center.

SCR is positioned as the base for fundamental research on future generations of semiconductors. It has a 3,000-square-meter clean room with Class 3 cleanliness on JIS standard (fewer than 1,000 particles smaller than 0.1 micron in 1 meter cubic, which is almost the same as ISO14644) and another 1,500 square meter clean room with Class 5 cleanliness (fewer than than 100,000 particles per 1 meter cubic). Activities at SCR will parallel those of Sematech in the U.S. and IMEC in Europe. SCR hopes to be the world's top R&D center within three years.

HALCA is a three-year project to develop production technology for the other three projects It is targeting technology development that can be economical even in small-scale production of 2,500 wafers a month and which cuts the energy consumption by 60 percent. This project will end by next March, the end of this fiscal year. The result and further development will be taken over by MIRAI and ASKA.

Asuka is a five-year project (2001 to 2005). Its mission is to establish a common platform for 65-nm node generations. Semiconductor Leading Edge Technologies, Inc. (Selete) is charged with device process development and Semiconductor Technology Academic Research Center is responsible for SoC design technology development.

Five research groups are working on high-K materials, low-K materials, new transistor structure and measurement, lithography and mask, and new circuit and system. The MIRAI team will make an intermediate report at the end of this fiscal year.

In the MIRAI project, 24 companies, including Intel and Samsung and 20 universities, are participants.

METI strongly supports these projects financially through its affiliated New Energy and Industrial Technology Development Organization, or NEDO. MIRAI is a 100 percent-supported project. HALCA is supported to the tune of about half of its total budget. Asuka is an all-industry project.

ASPLA intends to establish the technology that can be ported to volume production lines immediately.

A fab or fabs to introduce the production technology developed at ASPLA are in consideration as an extension of its activity. Some companies are developing an idea to construct a joint foundry that will be installed with technologies developed by ASPLA.

There are only two 300-mm fabs operating in Japan at present. One is the Hiroshima fab of Elpida Memory Inc. and the fab of Trecenti Technologies Inc., a subsidiary of Renesas Technology Corp.

Toshiba and Sony have announced 300-mm fab plans. Toshiba plans to begin operation of its Oita 300-mm fab in the first half of fiscal 2004. Sony plans to install the 300-mm line on the first floor of its Nagasaki Fab2 starting this fiscal year.

While NEC has expressed strong interest in the idea of the joint foundry (it provided the main part of the process technology that is used on the ASPLA test line), not all of ASPLA investors are in agreement.

For example, Matsushita, one of the primary investors, is considering building its own 300-mm fab based on ASPLA technology.

Still these R&D projects engender some criticism about overlapping projects and less than adequate cooperation.

"METI intends to review those projects within this year," said Hidetaka Fukuda, director of METI at the MIRAI annual achievement report meeting held in July.

The second SNCC is reviewing the activities of these projects and preparing a proposal to make those R&D projects work more efficiently. The proposal will be boiled down by next March, the end of this fiscal year, and those projects will be reformed according to the proposal.

Hitachi and Mitsubishi spun off their semiconductor businesses into Renesas Technology Corp., a joint venture company, which focuses on system-on-chip ICs and microcontrollers. The two parent companies announced the semiconductor merger in October last year, and the joint venture company started operation in April this year.

NEC spun off its semiconductor business and formed NEC Electronics Corp., which went public in July this year.

Elpida Memory Inc. was founded in 1999 by cobbling together Hitachi and NEC's DRAM business. The two companies shared ownership equally, but that made for slow decision making. Last November, NEC and Hitachi appointed a new president with allegiances to neither. Since then, Elpida has succeeded in raising funds from Intel and other investors in April to expand its 300-mm fab.

Fujitsu and Toshiba were also expected to merge their semiconductor operations eventually, but that never came to be. Instead, Toshiba is strengthening its semiconductor relations with Sony.

Through consolidations, reorganizations and the establishment of joint R&D initiatives, the stage has been set for Japan's re-emergence. Given all that, it would be hard to discount the possibility that Japan will regain the clout that it once wielded on the global semiconductor scene.

See related chart