A much different perspective emerges, however, when global industry technology investment is viewed through the more long-term, strategic prism of electronics and computer industry R&D.

Think of R&D in the U.S. high-tech sector, and names like Xerox Palo Alto Research Center, MIT Lincoln Lab, AT&T Bell Labs and IBM's Thomas J. Watson Research Center come to mind. Today, Bell Labs, T.J. Watson and Xerox Parc are mere shadows of their former selves, downsized by the realities of global business and technology challenges that now put the highest premium not on long-term research but on short-term product goals. Research today is a nearly extinct species of scientific inquiry in the electronics and computer sectors, but development is thriving. And it is being carried out by a massive diaspora of funds, projects and people peppered across the globe.

The migration "is not just about outsourcing for low labor rates," George W. Wang, director of IBM's China Development Lab in Beijing, said recently, noting that IBM has had a long history of globalizing R&D. "If it was just about low labor rates, we'd probably have R&D centers in places like Romania and the Philippines," Wang observed.

IBM's R&D strategy in China dates back 10 years, to when Wang transferred from IBM's T.J. Watson center in Yorktown Heights, N.Y., to set up IBM's China Research Lab in Beijing. More recently, IBM's focus has turned to more near-term technology and product development efforts in China, and Wang recently brought "D" of the R&D to IBM's China investment by establishing a software development lab in Beijing. Activity is centered on supporting the company's software business units. The lab also does development in Shanghai and Taiwan.

"China is the most important emerging market, and we need to have deep product skills" to support IBM's business there, Wang noted, outlining a comprehensive program of development projects that involve both IBM software developers and researchers from China's leading technical universities.

Software and technology development programs of this type are becoming common for a wide range of companies in the computer and electronics industry-not just in China but in India and other regions characterized by large, developing business and market opportunities.

Recently, Infineon announced plans to expand its Bangalore, India-based R&D center by adding 200 employees to the existing base of 250. The additions are expected to be made in the areas of software development, tool development and sales and marketing. The Munich-based company has a global work force of 32,900, out of which 6,100, spread across 27 centers in 10 countries, are engaged exclusively in R&D activities. The company spent roughly 1 billion euros on R&D (about $1.23 billion) from revenue of 6.15 billion euros for the fiscal year that ended in September.

The R&D center in Bangalore develops software for e-Farm, a pilot RF identification project that puts RFID technology to use tracking sheep.

Intel's Peter Liou: 'We follow the law of numbers.'

Access to markets and to the best and brightest engineering talent is a far more powerful driver of R&D globalization than mere labor rates, according to industry analysts. A case in point is Intel Corp's venture-capital arm. Known for funding promising hardware and software technology likely to spur demand for Intel products like the Pentium, Intel Capital has poured funding into Israel, which is hardly a low-cost labor market. Indeed, Intel Capital was Israel's top foreign corporate investor from 2000 to 2003, with 35 initial investments, according to the IVC Research Center, which tracks investment activity in Israel.

Like IBM, Intel also has a well-established, historical presence in China, and, increasingly, the focus is development rather than pure research.

"There's really not a clear line between research and development," Peter Liou, director of the corporate technology group of Intel China's Research Center, said during an interview in Beijing.

Intel also has close ties to China's technical and engineering universities. But even more important, Liou said, is being "close to the market from the standpoint of user needs and being close to the ecosystem where we can better enable our OEM and ODM partners."

If low labor rates are not the key attraction for locating R&D centers in places like India and China, what is the secret pulling power of these regions?

"It's the law of numbers," said Liou. "With 1.3 billion people in China alone, there's a very large pool of talent, and Intel gets to have its pick of the cream of the crop from this immense pool."

Intel also does a limited amount of microprocessor architecture, compiler and network and communications development work in China

But when it comes to Intel's most advanced development work, particularly in the areas of silicon layout and advanced processor design, "all of that is still done back in the U.S.," said Liou.

That's something the diaspora of R&D is unlikely to change anytime soon.