A NewCo to Revive the Past
It’s been well established by now that North America, once neck-and-neck with Japan for the title of largest producer of printed circuit boards, has long since ceded that space to China.
But there was a moment, several years ago, at an industry meeting where market research was broadly disseminated and debated, where Taiwan was
promoted – briefly – as the future.
Walt Custer, the charismatic pundit known for his informative and humorous presentations, was asked if North America was about to be eclipsed by the Pacific Rim. His reply was that he felt it was only a matter of time before Taiwan became the top-producing nation.
Dr. Hayao Nakahara, sitting in the audience, then stood and pointed out some basic facts:
- Taiwan had a population of around 23 million.
- China had a population of well over 1 billion.
I don’t recall the entirety of the exchange, but Naka’s point was that China would overwhelm its neighbors with abundant space for new campuses and a near-inexhaustible supply of lower-cost workers in what was then still very much a labor-intensive manufacturing operation. China, he correctly deduced, would surpass everyone: North America, Taiwan and Japan.
Fast forward to today. China is indeed the dominant player in bare boards, with roughly 60% of the world production taking place there. But Taiwan is No. 2, at $11.6 billion as of 2021. That’s slightly more than the Americas, Europe and Africa and the Middle East combined. (For the full accounting, see Naka’s annual NTI-100 list. https://www.pcdandf.com/pcdesign/index.php/editorial/menu-features/16699-the-unsinkable-unstoppable-pcb-market)
But why was Taiwan not only able to surpass the West in output, but maintain that lead through two decades? Why hasn’t the island-nation succumbed to the economics of demographics (low population, an educated workforce that would have ample opportunities in non-manufacturing sectors, an average age that rose 24% between 2000 and 2018) coupled with a relative scarcity of natural resources, particularly available land?
I asked a few experts on Taiwan’s bare board industry. The short answer: There’s no consensus. One did point to Taiwan’s government policies, which are conducive to maintaining key industries like semiconductors and PCBs. “Government is unified in supporting industry,” he said. A few others said the levels of automation in Taiwan far exceed those of North America. But could those factors alone be enough to allow Taiwan to continue to beat – easily – the West on price even after the shipping and other logistical costs are accounted for?
And if so, there’s a silver (or perhaps copper) lining.
Buoyed by government grants and the US Congress’ newfound passion for building things at home, a group of US-based companies are looking at closing the technical gap.
Under the auspices of the US Partnership for Assured Electronics (USPAE), these companies are tackling the issue of how to build substrate-like PCBs for next-generation semiconductors. In short, what good is $50 billion of new semiconductor fabs if the bare die must be shipped to the Pac Rim to be packaged?
“The concern is the US can’t do enabling technology,” said Joe O’Neil. “We need a highly automated process, with operations performed in clean-room environments, capable of low-mix, high-volume production.”
The short-term solution calls for setting up a service bureau to work with existing suppliers to produce the core technology, which would then be sent to a dedicated company (called NewCo) to make the buildup layers using a common process capable of 1/1 mil lines/spaces.
“That would get things started,” says O’Neil, “but won’t solve the industry base problem.”
The longer-term goal, he says, is the creation of a Center of Excellence for domestic PCB training. The outfit would develop the recipes, prove them out, and transfer that technology to industry, and train the workforces of the participating members. The objectives call for settling on a single process in order to accelerate the learning curve, so less time is spent on R&D and more on training.
The actual training goals are fairly modest: about 12 engineers every six months. But it’s the closest the US industry has come since the conception of the Interconnection Technology Research Institute to forming a manufacturing collaborative with an actual physical production space. (We say “conception,” because ITRI never realized its own plant.)
We can’t just wish a turnaround to happen. If North America wants to revive its past, it needs to invest in technology that will allow it to leapfrog its current trendline. Given the stakes, we applaud this effort to match ideas with action.