SAN FRANCISCO Aggressive predictions about the speed and prevalence of semiconductor manufacturing moving to China may be off base, according to a top executive at Advanced Semiconductor Engineering (ASE) Inc.
Tien Wu, president of ASE's U.S. subsidiary, ASE Inc., told an audience here at the Semicon West tradeshow Tuesday (July 12) that, while China is undoubtedly rapidly become a global semiconductor manufacturing power, bold predictions about the speed at which manufacturing will move to the country, such as Semiconductor Equipment and Materials International (SEMI)'s recent proclamation that China will build 20 new fabs between 2005 and 2008, are overstating the situation.
"I don't believe it," Wu said. "It is not a forgone conclusion that all manufacturing will move to China."
Wu argued that a certain percentage of semiconductor manufacturing would continue to be based in the U.S. and Europe because, he said, some percentage of chip making will always be local.
Wu pointed out that China's evolution as a semiconductor manufacturing power is completely different than any before it because China is moving directly to a fabless-foundry model, while the U.S., Japan, Taiwan and Korea could rely on the infrastructure created by integrated device manufacturers (IDMs). Wu said that he could trace the roots of fabless semiconductor companies back to IDMs in every other region of the world, and that predictions of the industry's growth in China underestimate the value of relying on the provided infrastructure.
"China needs to figure out a very different path than the U.S., Taiwan, Japan and other regions," Wu said. "They believe that with 500 design house and a bunch of foundries they can wage this campaign."
The evolution of Chinese semiconductor manufacturing faces challenges in the form of sustainability (in terms of the cost of raw materials), efficiency and the political stability of the region, Wu argued.
Wu said the lower cost manufacturing possibilities offered by China would not necessarily propel the country to total semiconductor industry dominance. Life sciences, defined by Wu to include replacements for human body parts and other innovations, will be the next major driver of the IC industry, he said, and the cost of devices will not be the most important factor to consumers.
"As applications get closer to the human body, human psychology plays a more important role in the buying decision," Wu said, noting that people tend to buy based on price for things that are not with them all of the time, such as ink jet printers, but care more about quality and style in products that they carry all of the time, such as cell phones. When it comes to items that actually reside within the human body, Wu said, people will want them to be powered by higher quality, more expensive chips.
"If you want to embrace high volumes, [establish manufacturing in] China," Wu said. "But you have to figure out what your value proposition is."
The size of China as a market for electronics may ultimately play a larger role in China's emergence in semiconductor manufacturing than lower costs, Wu said.
"Dell didn't go to China because they wanted to reduce costs," Wu said. "They want to sell a PC to every Chinese."