By David Lammers

Larry Sumney foresees more systems-level research, more involvement in the Middle East, and more 3D-related programs for Semiconductor Research Corp. (SRC).

Larry Sumney

The SRC is approaching its 30^th anniversary next year, with about 100 universities participating in its programs this year. Sumney, the SRC director for those three decades, had a distinguished career as a Naval Research physicist and ran the military’s billion-dollar Very High Speed Integrated Circuits (VHSIC) program before being named to head up the SRC in 1982.

Most of the SRC programs thus far have been aimed at “how to take CMOS to the limit,” Sumney said in an interview at the SRC’s TechCon event in Austin. Now, with so few companies financially able to invest in leading-edge device technology, the emphasis is changing. More research programs are aimed at system-level research, including circuit and system design. The new Multi Scale System Center, MuSyC, located at Berkeley and headed up by Prof. Jan Rabaey, is one example.

“Our members are becoming more interested in what we are calling ‘functions per unit volume,’” which includes stacking chips with TSV connections. “Scaling will continue, but it will become more and more expensive, and only a few companies can afford it,” Sumney said.

The Focus Centers, based at M.I.T, Berkeley, and other universities but involving larger networks of universities, are being reconfigured to adjust to these changing priorities. The emphasis is shifting to systems, chip stacking, and “adjacent spaces” such as biosensors and security. Sumney uses the word “recompete”  to describe how the SRC’s board obsoletes some Focus centers, puts out RFPs on new centers, and keeps the whole effort up-to-date.

“The Focus Centers are being redefined right now, but the result I am sure will be more emphasis on systems,” Sumney said.

Analog is another push. Texas Instruments, as part of its decision to have its foundry partners do much of the digital CMOS technology development, has shifted R&D toward analog. TI told the SRC board that it was less interested in funding the SRC’s programs. Instead of seeing TI go by the wayside, the SRC team put together the Texas Analog Center of Excellence, TxACE, based at the University of Texas at Dallas. Funding is shared equally by TI, the state of Texas, and the SRC. Prof. Kenneth O was enticed to move from Florida to Texas to head it up, and he in turn recruited several other analog design gurus.

While the SRC’s roots are in America, it is an international organization, with Tokyo Electron Ltd. and ATIC (the Abu Dhabi Technology Investment Corp.) as members. Earlier this week, seven research contracts were announced by the SRC, awarded to four universities in the United Arab Emirates (UAE), with funding coming from ATIC. Those professors and students will collaborate with a network of U.S. universities and research labs. Sumney said it is highly possible that a similar set of research efforts will be established with KAUST University in Saudi Arabia.

“We tend to think that the output of these programs is the research results,” Sumney said. “But for our members, they are equally interested in identifying the good students, people they can hire. That is where much of the return on investment is for our members.”

The SRC is establishing new efforts in solar energy, smart power, energy storage, and power management, among others. Over its 30 years, the SRC has evolved.

“To succeed, we need to be creative,” Sumney said.

By David Lammers

In 2006, I had an interesting interview with Jim Doran about cultural diversity and chip manufacturing. Doran, who held manufacturing executive positions during a long career at Intel, AMD, Spansion, and finally GlobalFoundries, made several points that bear a brief retelling:

“This industry is an interesting mix of business, science and people,” Doran noted. “We all deal with electrons that move around in the same way, but there is a huge difference in how companies get value out of their asset base — company to company,” he said, detailing several methods he found useful to bring out the best in people from different cultures.

For example: Spansion, where Doran worked at the time of the 2006 interview, needed to come up with common metrics to measure productivity at its fabs in Austin and northeastern Japan, such as line yield, cycle times, and labor productivity.

I asked Doran if Japanese engineers, imbued with a pride in things Made in Japan, were “stubborn.”

Doran replied that “it often does take a longer time to get our Japanese colleagues to accept a new or radical idea. Any culture has pride and wants to be successful, and wants to build on that success. So yes, their first choice tends to be resistance. The other side of that is, once you do get an idea accepted, there is going to be a march-to-the-goal-line kind of thing, in a very dedicated way. You have to spend more time up front building consensus in Japan, but once you do, by my observation, then you just have to get out of the way.”

Doran’s answers were accurate, insightful, and extremely credible, given his experience building and managing fabs around the world.

GlobalFoundries is not the only multi-cultural company in the semiconductor industry, of course, but it may be unique in that among its 11,000 employees it has an almost equal-parts mix of Americans (which involves several dozen different nationalities, for example, currently working at the Malta, N.Y. Fab 8), Germans at the Dresden Fab 1 complex, and the Singaporeans with their hard-won foundry management experience. And now we can add people from Abu Dhabi and the Middle East, with a likelihood that the eventual Abu Dhabi fab will be staffed with hundreds of engineers from India and the other nations in the sub-continent.

What other company is so decidedly multi-cultural in its ownership, management, and overall workforce?

At the GlobalFoundries technology event in Santa Clara, the multi-cultural theme kept popping up. Because managers from the former Chartered had experience running multi-product fabs, a team of ex-Chartered managers transferred to Dresden to infuse the multi-product culture into the German engineers there.

Rutger Wijburg

The senior executive at Dresden currently is Kay Chai Ang, in charge of 300mm fab operations at GlobalFoundries. GlobalFoundries recently hired an ex-NXP fab manager, Rutger Wijburg, as the new general manager of  the Dresden operation. Wijburg, a Dutchman who speaks several European languagues fluently, will report to Ang. Interim CEO Ajit Manocha said he is spending much of his time in Dresden to manage the ramp there.

The mix of cultures is bound to be a long-term positive for GlobalFoundries. It should not be overlooked that Singapore itself is a multi-cultural society, with the dominant Chinese co-existing profitably with Western expats, a large infusion of Malaysians who provide much of the fab labor, and people of Indian heritage.

The German workers bring a dedication to quality. Just as Germany’s car industry has done well at the high-performance end of Audi, Porsche, BMW, and Mercedes Benz, Germany’s semiconductor engineers bring a similar devotion to engineering excellence. For GlobalFoundries Dresden, that was expressed in the manufacture of AMD’s microprocessors, a high-performance product line.

The question facing the Dresden work force is whether it can increase its flexibility, ranging from the ability to make multiple and lower-cost products to an ability to work well with non-German engineers who might have a different view of how problems can be solved.

For the Singaporeans, creativity is a challenge. In Confucian cultures such as China and Japan, the social hierarchy is important. Coming up with new and innovative solutions to chip manufacturing challenges perhaps doesn’t come quite as naturally to the Singaporeans as it needs to.

Since I know very little about Abu Dhabi or the people there, I had best refrain from saying anything about the subject.

The cultural diversity of GlobalFoundries is its greatest asset. Developing an openness to the unique characteristics shared by people from other cultures is something that doesn’t come naturally to all problem solvers. But it must be a top priority for this global foundry as it searches for a new CEO.