Can the U.S. Keep Up with Asia in the Chip Wars?

Sometime before 2007, Paul Otellini bet against his instincts.

The then-CEO of Intel Corp. received an offer from Apple’s Steve Jobs: Could Intel make chips for its forthcoming innovation, the iPhone? Otellini passed on it.

It was a decision Otellini would later regret.¹ Apple went with Taiwan Semiconductor Manufacturing Co. (TSMC) for its smartphone chips instead.

“The lesson I took away from that was, while we like to speak with data around here, so many times in my career I’ve ended up making decisions with my gut,” Otellini told The Atlantic in 2013. “And I should have followed my gut. My gut told me to say yes.”

Smartphones now occupy 30% of the demand for semiconductors, with Apple as the largest customer. Nearly all production for smartphone chips occurs in Asia.²

Missing out on smartphones isn’t the only reason the U.S. trails East Asia in the chipmaking business that it once dominated. Blame also goes to higher labor and capital costs in the U.S. that contributed to chip production migrating to Asian competitors. The U.S. currently manufactures 12% of chips globally, down from 37% in 1990.³

The U.S. recognizes it needs to make up ground from an economic perspective and as a national security matter. In 2022, Congress passed the CHIPS Act  to catch up with Asia, particularly Taiwan and South Korea, which along with China and Japan, make 75% to 80% of the world’s chips.⁴

The measure makes $53 billion available for domestic chip manufacturing, research and development and workforce training over the next five years.

The CHIPS Act allows the U.S. to reclaim territory in the semiconductor industry. But the legislation will only make meaningful strides globally with further investment from Congress.

Still, $53 billion matters. We see winners and losers emerging from U.S. investment in the industry. And somewhat ironically, Intel may emerge as the biggest winner.

Why Does the CHIPS Act Matter?

The COVID-19 pandemic and its ensuing government lockdowns and supply chain disruptions brought into focus American weaknesses in semiconductor manufacturing. If U.S. consumers are startled at the sticker prices for new cars, part of the blame can go to the bottleneck of trailing-edge semiconductors coming out of Taiwan during the pandemic.

A Harvard Business Review study examined how even a short, 10-day disruption of a semiconductor fabrication facility — or a “fab” — would cause a chain reaction affecting the entire supply chain for almost a year.⁵

The Center for Security and Emerging Technology described the complexity of the half-trillion-dollar supply chain in a 2021 paper:

The U.S. remains a global leader in chip research, but most chip manufacturing occurs in East Asia.

The National Security Commission on Artificial Intelligence reported that dependency on semiconductor imports, especially from Taiwan, exposes the U.S. to economic and national security vulnerabilities.

“If a potential adversary bests the United States in semiconductors over the long term or suddenly cuts off U.S. access to cutting-edge chips entirely, it could gain the upper hand in every domain of warfare,” the NSCAI said in asserting the U.S. could regain meaningful capacity in semiconductor manufacturing.⁷

China might be that one potential adversary. China, like the U.S., wants to make up ground in the semiconductor race. It has spent $150 billion since 2014 to expand its industry and announced aggressive plans⁸ — roughly another $145 billion worth of incentive packages for its domestic chip industry⁹ — with little regard for financial return on its investment.

China, which manufactured no semiconductors in 1990, wants to occupy 24% of the world’s share by 2023.¹⁰

We believe China’s investments pair with another risk that could complicate the U.S. investments in chips: Tension between China and Taiwan.

Geopolitical tensions between them have continually increased since at least the last decade. The potential for a Chinese invasion of Taiwan worries Western countries for several reasons, including the potential for a military standoff with China.

China taking control of Taiwan’s robust semiconductor industry is another potential issue.

We don’t know what China will do, and experts offer a range of scenarios.

Matthew Pottinger, a deputy national security advisor during the Trump administration, recently told The Washington Post that China had made various militaristic investments that have received scant attention. He believes 70-year-old Chinese President Xi Jinping views a Taiwan takeover as “his legacy achievement.”¹¹

If such a scenario came to pass alongside China’s aggressive investments in semiconductors, we think it would further diminish the U.S. position and interests in the so-called chip wars. It would also reveal that investments in the CHIPS Act were insufficient to compete with China.

How We Got Here

It wasn’t always this way. The U.S. and Europe held strong positions in the semiconductor industry from the 1960s to the 1990s. In 1990, the U.S. and Europe combined to make 81% of the world’s semiconductors, and Japan made the rest. Then China, Taiwan and South Korea nudged their way into chipmaking and have been increasing their prowess ever since.

The script has flipped: Asian countries now make 75% to 80% of the world’s chips, while the U.S. and Europe wish they made more.

Labor costs in the U.S. played a significant role in driving manufacturing to Asia. Goldman Sachs reports that a semiconductor engineer in the U.S. draws a salary nearly four times what an equivalent would make in Taiwan. And after manufacturing overhead, raw materials and consumables enter the equation, the total operational cost to run a fab in the U.S. would be 38% higher than in Taiwan.¹²

Morris Chang, TSMC’s founder, told the Central News Agency that it cost his company 50% more to make a chip in the U.S. than it does in Taiwan, reflecting his broader skepticism about the U.S. effort to gain ground in chipmaking.¹³

The U.S. isn’t alone in its chipmaking aspirations. As Congress passed the CHIPS Act, the EU announced a plan to invest up to $50 billion in chips to reach a 20% global market share by 2030. China has 10 new fab projects. And Korea has the K-Semiconductor Belt, a strategy to invest $451 billion to forge the world’s largest semiconductor supply chain by 2030.¹⁴

CHIPS Act Winners and Losers

We believe Intel will likely receive the largest proportion of CHIPS Act spending. Piper Sandler shares that view, estimating that Intel could fetch nearly $20 billion of the $39 billion made available from the bill to manufacture chips¹⁵ (Bank of America puts it at $10 billion to $15 billion).¹⁶

Intel announced plans to build two new leading-edge fabs in Ohio at an initial investment of more than $20 billion.¹⁷ Intel announced plans to build two more fabs in Arizona, also at a $20 billion investment.¹⁸ Intel’s long-term investments mean the company, at least for the short term, plunged into negative free cash flow territory, a contrast from its cash-rich past.¹⁹

Whether Intel can catch up to TSMC on the technological front is another question. From our perspective, Intel didn’t do enough to embrace advanced lithography, which put the company behind TSMC technologically by two or three years.

But we also see an opportunity for Intel if it can develop a viable foundry — a manufacturing facility that makes chips at the request of other companies — in the U.S. We believe that would give companies like NVIDIA  and Broadcom alternatives to Taiwan, where chip production commands a steep markup.

Texas Instruments, a leading manufacturer of analog chips, spent $3.3 billion in capital expenditures and expects to continue making investments to shore up domestic production of trailing-edge chips, which suffered from pandemic-era bottlenecks.²⁰

We see semiconductor capital equipment companies (semicap) — vendors that provide tools and equipment for chipmakers — emerging as indirect winners from increased chip production. The semicap industry has few competitors and should benefit significantly from increased chip production. Companies like ASML N.V., Applied Materials and KLA Corp.

The CHIPS Act could pose problems for TSMC. The Taiwan firm plans to build two chip factories in Arizona at an overall investment of $40 billion. TSMC wants a reported $15 billion worth of tax credits and grants from the CHIPS Act but objects to some of the federal government’s strings attached to the legislation.

If Intel succeeds in developing a foundry in the U.S., we believe TSMC could lose some market share.

CHIPS Act: A Promising Start for Critical U.S. Investment

Overall, we believe the CHIPS Act made sense for the U.S. It’s a start at reinvesting in an industry critical to the economy and national security. The pandemic had few silver linings; maybe one was supply chain disruptions clarifying why semiconductors are essential and why the U.S. should make more of them.

In our view, the CHIPS Act shouldn’t be interpreted as a solution for the U.S. chip industry but as a starting point for continued investment.