PARIS -- The Europeans seem to be getting their act together again in chip R&D. Their Medea program got off to a slow start when national governments balked at coming up with the funding in the initial two-year phase of the initiative. But now the $2.2 billion program is finally moving ahead.
Primary goal is to accelerate "precompetitive" R&D work in order to make the European semiconductor industry more competitive globally in targeted chip applications. At stake not only is Europe's global presence in world markets, but also the jelling of Pan-European cooperation in IC-related R&D.
First phase of Medea wound up at the end of last year. Five chip-processing technology efforts were completed. But a multimedia project was killed by the companies involved in this work, a chip-packaging program was transferred to another European cooperative program, and an analog/mixed-signal design initiative was halted by consortium. The remaining 31 projects from Phase I now have been transferred and combined with 13 other "labeled" projects, which will all get government support under the Medea program until the end of 2000.
While it now appears that Medea finally is on its way to delivering results on several of its IC application targets, the program was slowed in its first year by an old European problem - lack of government funds (see story in the April 1998 publication).
Foot-dragging by some country governments started slowing the funding of R&D projects in 1997, but better economic conditions in Europe and progress in setting up a single European currency -- the euro -- has helped to get momentum going again in Medea (which stands for Microelectronics Development for European Applications).
About 40% of the total funding (2 billion euros) was spent on Medea during Phase I, and now it looks like the remaining 60% will be pumped into the program in its final two years (1999-2000), according to Grard Matheron, office director for the cooperative, which falls under the aegis of Eureka, a Pan-European research group supported by 20 European countries.
Medea now has nearly 130 partners, including companies, universities, and public laboratories, from 11 European countries. One factor that changed recently is that the number of small companies participating in Phase II has gone up by nearly 12%. "This is a good sign, showing not only that large companies can cooperate but also the smaller ones," says Matheron. "Many of these small companies were spinoffs from earlier programs and the Jessi projects," he adds, referring to the $2.7 billion Joint European Submicron Silicon Initiative completed by the end of 1996.
Medea's charter was to take up where Jessi left off. Some industry observers believe the real payoff from the eight-year Jessi program was in developing Pan-European cooperation, across country borders, cultures, and individual company strategies.
"Probably the first four years of Jessi was spent trying to figure out how to make it work," says analyst Malcolm G. Penn, president of Future Horizons in Kent, England. "Then everything seemed to come together in the last two or three years of Jessi," he says, "and now a large number of people involved in that program are part of Medea, which looks like it's now working well."
Progress in IC processing technologies ended up coming much faster in the initial stage of Medea than did advancements in design automation or chip applications, admits Matheron. "It's easier for the IC companies to predict the future of semiconductor technology because we have roadmaps," he adds, "and new generations are now coming in just two years."
In fact, the acceleration of chip technology led the Medea staff to ask participating companies to stop developing analog/mixed-signal design tools for 0.5- and 0.35-micron processes and reset the goals for 0.25- and 0.18-micron feature sizes during phase II.
A dozen of Medea's 44 projects currently are in process technology and tool development. One such effort in ion projection lithography aims to create a non-optical exposure technique for 0.07- or 0.05-micron feature sizes that will needed later in the next decade. The rest of Medea's Phase II projects are directed at putting advanced IC technologies to work in three main application areas: multimedia, communications, and automotive/traffic systems.
The bulk of the efforts in Medea's Phase I were focused on defining system-on-chip architectures and chip-set partitioning or definitions. "The actual silicon for many of these efforts is due this year or next," Matheron says. For many observers, delivery of chip application results ultimately will determine the success of the overall program.
As Medea partners work to complete those chip application projects, they also are now in the early stages of discussing the possibly of continuing Pan-European efforts in "precompetitive" R&D. "Our top priority today is to have each of the projects reach their goals and deliver results," Matheron says, "but we also are starting to think about 'phase III' or some other program for cooperation after the year 2000."
European R&D efforts also most likely will leverage some of the work already being done in other consortiums around the world, especially in U.S.-based Sematech which opened up some of its R&D last year to international partners. Europe's Philips, Siemens, and STMicroelectronics are now members of International Sematech.
"Cooperation among the worldwide players is really a must and we continue to discuss possible work with Sematech," Matheron says. But repeated efforts to cooperate with Sematech were unfruitful during the Jessi program in the early 1990s. "Cooperation is not so easy because our targets are quiet different," Matheron says. "Sematech is focused on manufacturing and process equipment while Medea is looking at the interaction between IC and system manufacturers." -- J. Robert Lineback