Nikolajs Strigins/Dreamstime
Electronic manufacturing with checkered flag

The Semiconductor Race is On!

March 23, 2022
Designing chips isn’t the issue—manufacturing is. Finding new capacity is an imperative with geopolitical implications.

At a Glance:

  • A confluence of events along the half-trillion-dollar global chip supply chain is behind the semiconductor shortage.
  • The biggest threat to the semiconductor industry in the United States is that semiconductor manufacturing has suffered a steady decline for decades.
  • In the U.S., designing chips is really not a problem—manufacturing is the critical issue, according to Brian Matas, VP, market research, IC Insights.
  • The concern for U.S. policymakers is that the semiconductor manufacturing is concentrated in the Asia Pacific region, which leaves America exposed to geopolitical risk.

When supply chain vulnerabilities resulted in breakdowns in shipping and logistics systems and shortages in personal protective equipment, a boom in demand erupted for cloud services and home electronics. And then came the pandemic, further straining both manufacturing and distribution.

A confluence of events along the half-trillion-dollar global chip supply chain unfolded throughout the pandemic. Manufacturers dealt with issues ranging from trade issues between China in the U.S. to global output and logistical challenges. Then there were the human and man-made disasters, including earthquakes and a fire in a major semiconductor plant in Japan.

In early 2020, vehicle sales plummeted, forcing automakers to cut chip orders. By the time there were signs of recovery in the latter half of the year, the semiconductor industry had already shifted production lines to meet demand for such product categories as 5G smartphones, laptop computers and webcams.

The ripple effect extended beyond semiconductors, said Peggy Carrieres, vice president of Sales Enablement and Supplier Development, Avnet, a global semiconductor distributor with headquarters in Phoenix, Ariz.

“Semiconductors are historically a cyclical business, so we usually expect to see a market cycle ebb and flow about every four years,” said Carrieres. At the cusp of the pandemic in 2019, an industry forecast from the World Semiconductor Trade Statistics (WSTS) organization projected annual global sales would slump 12.8%. Markets were expected to rebound with moderate growth and forecast worldwide semiconductor market growth to rise to 25.6 % in 2021 from 6.8% in 2020. The boost, noted WSTS analysts, is the biggest since a 31.8% increase in 2010.

Chip Design Thrives

A more complete view of the chip shortage traces its roots to long-term forces. The semiconductor industry has evolved into a highly specialized industry racked by intense competitive pressure. From its origins in Silicon Valley in the 1970s, semiconductors have roughly doubled in circuit density every two years, when Gordon Moore, co-founder of Fairchild Semiconductor and former CEO of Intel, made the historical trend known as Moore’s Law. Simplistically described, the moniker means one can cram more functionality into a smaller chip and that more chips per wafer plummets the cost per unit.

The biggest threat to the semiconductor industry in the United States is that semiconductor manufacturing has suffered a steady decline for decades. Fabrication capacity has fallen to about 12% in 2020 from about 37% in 1990, according to a report by the Semiconductor Industry Association (SIA) and the Boston Consulting Group. These statistics tell us that few semiconductors are actually made in America. Many major semi firms—Nvidia, Qualcomm and AMD—design and test chips in-house, but outsource production to foreign foundries.

The concern for U.S. policymakers is that the semiconductor manufacturing is concentrated in the Asia Pacific region, which leaves America exposed to geopolitical risk, trade restrictions and further supply chain disruptions. “We would have better control if [semiconductors] can be locally sourced,” said Brian Matas, vice president, market research, IC Insights.

Local production had proven to be cost-prohibitive for most private companies and makes clear why there’s no short-term solution for bringing semiconductor manufacturing back to the U.S. The 10-year total cost of ownership of a new fab in the U.S. can run 30% to 50% higher than competing countries, according to SIA data.

“What many in the United States think about is that one can throw $50 billion toward the infrastructure in the semiconductor industry,” said Matas. “That’s a nice plan, but there’s no immediate fix. Rather, to become a manufacturing power, we need a long-term plan and to spend significantly on the semiconductor industry in the country.”

China Wants to Take Charge

Advanced manufacturing capabilities, complex equipment and the capital outlay needed to compete globally have in recent years rested semiconductor manufacturing in the hands of a few companies and countries, said Matas. This is largely owing to government incentives and subsidies afforded to chipmaking companies in that region.

Backed by government financing, China has been able to bolster industry collaboration and expansion for the country’s fabrication capacity. Still, its role in the semiconductor manufacturing remains limited. Integrated circuit (IC) products produced by Chinese firms are generally less technologically advanced than those produced by non-Chinese fabs. EurasiaGroup, a political risk advisory, reported that while Beijing will continue to pour funding into the sector, manufacturers there will be challenged to penetrate the upper tier of global manufacturers that produce chips at nodes below 7-10 nm for the foreseeable future. That’s because growing costs hamper their ability to install the semiconductor manufacturing equipment and capabilities needed to produce volumes at process nodes below 10 nm.

In general terms, “nodes” refer to the generation of chips for a particular technology. The complexities of shrinking (remember Moore’s Law), expertise and high capital cost have propelled only three companies to the forefront. Intel in the U.S., Samsung in South Korea and TSMC in Taiwan can fabricate integrated circuits at the process nodes of 7 nm and are charging ahead to transition to 5 nm and 3 nm capacity by the mid-2020s. China’s largest semiconductor foundry, Semiconductor Manufacturing International Corporation (SMIC), reportedly lags by about three years.

Policy and Protectionism  

The Asia-Pacific monopoly has bolstered concerns about the long-term security of supply chains to the extent that it has become a matter of national importance to the United States. The Biden Administration’s $2 trillion infrastructure package includes a proposal of $50 billion in funding earmarked for such initiatives as a public-private consortium to build and operate a National Semiconductor Technology Center for conducting research and prototyping.

To compete with foreign competition, an aggressive policy framework is being enacted via the CHIPS for America Act and is buoyed by federal legislation (FABS Act), which promises incentives that could infuse $52 billion into semiconductor manufacturing, research and development, design and manufacturing. BCG analysts estimate that the incentive program can strengthen competitiveness by bringing in 19 new fabs (fabrication labs) online. The gains would accrue to a share of global installed capacity of about 13% by 2030 and increase local capacity by 57%, which reverses the downward trend of the past three decades.

As the global shortage continues, semiconductor fab investments are materializing around the world. According to industry association SEMI, 29 new fabs started construction in 2021 and 2022, but many won’t start installing equipment until 2023. It takes up to two years after ground is broken to reach that phase, though some could begin equipping as soon as the first half of next year, noted a SEMI press note.

In the U.S., designing chips is really not a problem—manufacturing is the critical issue, Matas noted. With support from government, he said the numbers are expected to climb. In his view, Samsung’s manufacturing fab in Taylor, Tex. is proof of that. The $17 billion investment, including buildings and machinery and equipment, is Samsung's largest-ever investment in the U.S.

“There’s a lot around the corner to keep the semiconductor industry active and vibrant,” said Matas. “And that’s just one reason it’s important to decrease reliance on foreign production and bring some of the manufacturing back.”

Voice your opinion!

To join the conversation, and become an exclusive member of Machine Design, create an account today!