Trump’s “Reciprocal Tariff” Strategy and Section 232: What It Means for TSMC, Intel, and the Future of the Global Semiconductor Race

Since returning to office in February, President Trump has introduced new initiatives almost daily. The latest rumor suggests that the U.S. government is planning to invest directly in Intel. Previously, Trump reportedly expressed dissatisfaction with Intel’s slow progress and was said to be considering the removal of its current CEO, Lip-Bu Tan. In parallel, he imposed tariffs on all foreign manufacturers exporting to the U.S., and more recently demanded that Nvidia pay 15% of its revenue from H20 chip sales to China to the U.S. government. These moves could have far-reaching consequences for the global tech and semiconductor industries and merit further scrutiny.

I believe the U.S. government’s plan to invest in Intel is a positive signal for the company. Intel is currently burning through cash at a rapid pace and can no longer afford to invest in advanced process development or production capacity. Government funding and support would certainly help stabilize the company and support its transformation.

However, even if the U.S. government increases its investment in Intel, it’s unclear how much funding it can actually provide. More importantly, what other policy moves might follow?

I believe it’s entirely reasonable for the U.S. government to support its own companies—many countries are doing the same. In fact, beyond providing funding, the government will likely introduce favorable policies to benefit Intel, which seems inevitable. However, before implementing such measures, some fundamental principles must be respected. Otherwise, it’s uncertain whether Intel can be successfully revived.

Even if Intel receives government funding, it is critical that the company maintains operational independence and avoids excessive government interference in its internal management. While some Asian countries frequently pursue state-led industrial upgrading policies, American corporate competitiveness has traditionally been rooted in market-driven mechanisms and private-sector innovation. If President Trump and his team believe government intervention is the solution, they may be headed in the wrong direction—and unlikely to achieve the desired results.

Recent market speculation suggests that, in its effort to revive Intel, the U.S. government might attempt to involve TSMC and pressure the company to transfer technology. However, such an approach would be highly inappropriate. TSMC and Intel differ significantly in their technological foundations, corporate cultures, and talent structures. They are not naturally complementary, and forcing the two companies together would likely produce no synergy. Instead, it could lead to a lose-lose outcome for both the U.S. and Taiwan—something the Trump administration should avoid.

Intel’s future will hinge on its ability to deliver advanced process technologies. But in the longer term, the best strategic path for the company is to spin off its manufacturing division and merge it with GlobalFoundries.

Two U.S. Foundry Leaders Could Merge into a Comprehensive Chipmaking Powerhouse

Intel and GlobalFoundries are both American companies that share similar corporate cultures and business DNA, making integration more effective. Their technologies and product lines are highly complementary: Intel focuses on advanced nodes, while GlobalFoundries has a well-established presence and customer base in the mature 12nm-and-above segment. A merger between the two could create a fully integrated foundry powerhouse.

GlobalFoundries’ most advanced process technology is at the 12nm node. While it isn’t cutting-edge, it is sufficient to meet the needs of many mature and specialized applications. Its customer base includes IC design firms and IDM companies across sectors such as automotive, industrial, communications, and consumer electronics. Key clients include AMD, Qualcomm, Broadcom, and MediaTek—relationships that would be highly valuable for Intel as it attempts to enter the dedicated foundry business.

Historically, Intel has lacked the service-oriented DNA required for the foundry business. If it continues to operate under its current structure—where product development and manufacturing are tightly integrated—the foundry effort is bound to fail. These are fundamentally different businesses with distinct cultures. Forcing two types of talent and operations into a single organization will inevitably lead to internal conflicts and erode customer trust.

If Intel were to spin off its manufacturing division and merge it with GlobalFoundries, the outcome could be dramatically different. Intel would gain the opportunity to learn how to truly serve customers, adopt a more humble approach, and earn the trust and business of clients.

Craig Barrett, former chairman and CEO of Intel, recently wrote an op-ed criticizing Lip-Bu Tan’s claim that Intel won’t invest in the 14A advanced node without customer orders, calling it a “joke.” Barrett stated that Intel now needs $40 billion in funding—an amount that the U.S. government is unlikely to fully provide, as it is comparable to the entire CHIPS Act budget. He urged eight major U.S. tech firms, including Nvidia, Apple, and Google, to each invest $5 billion in Intel to help build out its technology and capacity. Only with such support, he argued, can Intel compete with TSMC and maintain America’s competitiveness in semiconductor manufacturing.

Belett’s proposal may seem logical from a U.S.-centric perspective, but it is not necessarily practical within the global semiconductor ecosystem. At present, the U.S. significantly lags behind Asia in semiconductor manufacturing capabilities, and it is estimated that it would take at least 10 to 15 years to catch up through reinvestment. Belett’s idea—to have eight major tech companies collectively invest $40 billion to support Intel—does not directly address the urgent issue of foundry capacity shortages faced by the industry today. What companies need right now is stable and immediate manufacturing capacity and technical support—not to wait for Intel to transform and regain competitiveness years down the road. While the proposal has strategic intent, it is unlikely to garner broad support across the industry.

I’ve pointed out many times that the U.S. has already fallen behind in semiconductor manufacturing, which has shifted to Asia—specifically Taiwan, South Korea, and China. The critical gap lies in business model innovation. Yet Barrett still believes Intel can handle both chip design and foundry operations simultaneously, which is completely unrealistic. If you asked the eight companies he mentioned, it’s unlikely any of them would agree. For the U.S. to regain its manufacturing strength, it must undergo a fundamental mindset shift—especially abandoning Intel’s outdated practices. Otherwise, failure is inevitable.

In summary, my view is this: If Intel merges with GlobalFoundries, development of advanced process technologies should be led by Intel personnel, but the overall leadership of the company should be handed to GlobalFoundries. At the very least, the CEO should come from the GlobalFoundries side. The company has decades of experience in the foundry business, understands the industry’s dynamics, and knows how to serve customers. While it may not match the capabilities of TSMC or Samsung, it is deeply familiar with how this industry operates. In the foundry world, the essence of competitiveness lies in being humble enough to bend low before the customer.

The New Battleground of Hidden Tariffs: How the U.S. Profits from AI Chip Exports to China

In addition, Nvidia and AMD are now required to pay 15% of their revenue from chip sales to China to the U.S. government. This goes beyond Trump’s previous tariff measures and represents a new, far-reaching policy that could have significant implications for the industry’s future. It’s an issue that deserves closer examination.

The requirement to pay 15% of chip sales revenue from China to the U.S. government was recently proposed by the Treasury Secretary and has been confirmed by both Nvidia and AMD. This payment is part of a special arrangement in exchange for U.S. export licenses. Such a model could become a precedent for future cross-sector applications beyond traditional tariffs.

Fundamentally, this “revenue-for-license” model represents a more flexible and maneuverable form of indirect control compared to traditional tariffs. It can be viewed as a new kind of economic security tool. If widely adopted, it could reshape cost structures and compliance standards across global supply chains—posing particularly significant challenges for multinational tech and manufacturing companies that rely heavily on the Chinese market.

Furthermore, regarding the tariff policy initiated by the Trump administration, there is a strong likelihood that U.S. import tariffs in the range of 15% to 20% will become a new global norm in the long term. Since Trump set this precedent, it is unlikely that future presidents will reverse the policy. These tariffs are expected to generate hundreds of billions of dollars in annual revenue for the U.S. government, making them a critical pillar of fiscal stability.

Therefore, the impact of tariffs will be long-lasting. But what does a 15% to 20% tariff really mean? What major changes might it trigger in the future?

First, I believe one of the simplest principles is this: any company seeking to enter the U.S. market must reassess its competitiveness if its product gross margin is below 20%. Products—or even entire industries—with margins consistently below this threshold will struggle to compete and survive in the upcoming tariff battles. High-margin companies will inevitably outperform low-margin ones, and businesses with thin margins are unlikely to endure.

Therefore, if an industry’s average gross margin falls below 20%, it is in any country’s strategic interest to consider relocating that industry to regions with lower production costs and more favorable conditions. This is especially important in light of the United States’ new tariff strategy, which will reshape global competitiveness. If relocation still fails to improve margins, then exiting the business entirely should be a serious consideration.

The most important takeaway here is that companies must undergo transformation and upgrade, breaking free from low-margin industries and business models. This is a critical challenge that every nation and enterprise must confront. The principle is simple, yet brutally unforgiving—because it represents the most fundamental rule of survival in future global competition.

Taiwan’s Perspective on Small Manufacturing Nations: Diversifying Markets in the New Tariff Era

Finally, with a 15–20% tariff imposed on goods sold to the United States, companies must look to other countries and regions as key areas for risk diversification.

Take Taiwan as an example: Last year, Acer founder Stan Shih proposed that Taiwan should focus on what he called the “third world market”—meaning markets outside of the U.S. and China, including Europe and Japan. This combined market is potentially even larger than either the U.S. or China alone.

Today, Taiwan stands at the center of the global semiconductor and computer industries, and in the age of AI, these non-U.S./China markets will increasingly rely on Taiwan.

Every country with a thriving manufacturing sector must find diversification and growth strategies that suit its own national conditions.

Stan Shih has pointed out that non-U.S. and non-China markets rely heavily on the integration of hardware and software across various application domains—an area where Taiwan holds a clear advantage. Whether it’s core technologies like semiconductors and IoT, or application areas such as smart healthcare, smart cities, smart transportation, and smart agriculture, these are strategic entry points where Taiwan can exert meaningful influence.

While Taiwan faces inherent limitations due to its small domestic market, access to markets remains critical for innovation-driven businesses. Acer’s experience illustrates this well—despite strong sales performance in many countries, gaining a foothold in the U.S. and China proved far more challenging. This is a common experience for many Taiwanese firms. As such, the ability to penetrate non-U.S. and non-China markets will be a crucial factor in determining the success of future transformation and global expansion strategies.