(Re)Shoring Up Semiconductor Manufacturing

By Don Scansen, EETimes (March 22, 2021)

Sometimes history can be difficult to decipher. But the reasons why manufacturing moved out of the United States are blatantly obvious — cheap labor and reduced regulation.

Even though the analysis might be easy, history can still be difficult once those old mistakes start becoming clearer.

Sometimes recovering from mistakes is not possible. Will US domestic semiconductor manufacturing recover in a meaningful way? What factors might drive that success story?

Answering those questions seems a little ambitious, even for a blog post, but let’s attempt a few.

What’s changing?

Offshoring’s main selling point always was and is reduced labor costs. But as those cheap labor pools become upwardly mobile, demands for better living standards tended to exert increasing pressure on the original planning calculations.

Some jurisdictions, particularly while in the developing phase, offer relaxed regulations. But as countries become more affluent, they naturally want to take better care of their own backyard. Whether through government regulation or corporate ethics, increasing respect for the natural environment starts to cost more.

As semiconductor manufacturing plant costs move into eleven figure range, it is not a simple case of just moving to the latest frontier of cheap labor. The Chinese experience in wafer foundries is proof of that. Then there are bureaucratic challenges to setting up in China along with known risks to intellectual property.

Offshoring calculations are changing. The Chinese labor rate gap is closing. Thanks to explosive Chinese economic growth, real wages have increased by a factor of eight for urban Chinese (since 2000 as reported in Forbes). The balance is clearly shifting.

My view is that offshoring was quickly rationalized by viewing certain technologies as commodities. High volume components for the consumer market is not a matter of national security. Surely a DRAM mega fab or a custom foundry cranking out massive volumes of cheap computer chipsets or even modern cell phone processors is not strategic

To read the full article, click here