U.S. Government Takes 10%+5% Stake in Intel; Trump Administration Drives Spin-Off Plan — Who Would Bring the Greatest Synergy?

On August 22, 2025, the U.S. government announced it would acquire nearly 10% of Intel’s shares, with an additional clause allowing it to purchase another 5% in the future. This signaled that Intel may eventually spin off its manufacturing division. The move is expected to have major implications for the global semiconductor industry.

According to information disclosed by Intel and the U.S. government, the government converted $8.9 billion originally earmarked as subsidies into equity investment, becoming Intel’s largest single shareholder. The subscription price was $20.47 per share—nearly 20% below market value—bringing its stake to 9.9%.

The Trump administration stated that the investment allows U.S. taxpayers to buy into Intel at below-market prices, while sharing in the company’s long-term success alongside existing shareholders.

Intel, for its part, emphasized that the U.S. government’s investment is a “passive investment,” without board participation or access to privileged information.

The most significant meaning of this deal is that Washington has strengthened its grip on semiconductor manufacturing. What began as subsidies under the CHIPS Act has now shifted to direct equity and warrants, showing that the U.S. government is no longer just a supporter but a direct participant in shaping the industry structure.

Key Clause: What Lies Behind the 51% Threshold

In addition, the agreement includes a crucial detail that has largely gone unnoticed.

The deal grants the U.S. government a warrant to purchase an additional 5% of Intel’s common stock at $20 per share, but only if Intel’s ownership in its foundry business falls below 51%. In other words, if Intel sells more than 49% of its foundry operations, Washington can exercise this right.

So what does this specifically worded clause mean?

I believe it sends an extremely important signal that can be interpreted in at least two ways.

First, a “foundry spin-off” has become one of Intel’s strategic options, likely discussed during negotiations between the Trump administration and Intel. Washington also secured a lever to push forward such a restructuring in the future.

The warrant’s condition is explicit: if Intel’s control of the foundry business drops below 51%, the U.S. government can exercise its additional rights. The purpose is to ensure that even after design and manufacturing are separated, Washington can strengthen its control over Intel and play a leading role in shaping the spun-off manufacturing division, thereby securing domestic chipmaking capacity in the U.S.

Moreover, this indicates that spinning off the foundry business has become a concrete direction for Intel. The 51% threshold suggests that external capital will likely be brought in, paving the way for an independent company.

It is clear that President Trump, by publicly calling for then-Intel director Lip-Bu Tan to step down, was creating negotiating pressure. Later, when he summoned Tan to the White House for talks, it was highly likely that he explicitly hinted Intel should move toward a spin-off.

This negotiation once again showcased Trump’s deal-making style and revealed his clear intent to take the lead in steering Intel’s restructuring.

The Ideal Partner: GlobalFoundries

However, whether Intel will actually proceed with a spin-off remains unclear. The company has not made any public statements, likely because the board of directors has yet to reach a conclusion.

In addition, Intel’s old guard has consistently opposed a spin-off. For example, former chairman Craig Barrett has repeatedly contributed opinion pieces in the media, while former CEO Pat Gelsinger has spoken out multiple times in support. They strongly argue that Intel should maintain an integrated model of design and manufacturing in order to compete with TSMC and Samsung Electronics.

Therefore, whether to spin off Intel remains controversial in U.S. society. However, the Trump administration now appears to have made up its mind to push forward with Intel’s future spin-off plan.

Looking ahead, how will the spin-off be carried out? Which partners will be brought in? Which shareholders or funds will be invited to participate? All of these are critical factors that will shape the industry’s future development.

My view remains unchanged. If Intel spins off its manufacturing division, the most suitable partner would be GlobalFoundries. A partnership or merger between the two would face the least resistance, generate the greatest benefits, and deliver the strongest synergy. This would not only help the U.S. rebuild its semiconductor manufacturing strength, but also provide a positive direction for global industry development.

GlobalFoundries originated from AMD’s manufacturing division. It merged with Singapore’s Chartered Semiconductor in 2009 and acquired IBM’s Microelectronics division in 2015, evolving into a global foundry with operations spanning the U.S., Germany, and Singapore. While it incorporates some Asian and European elements, it remains one of the few specialized foundry companies whose core organization and culture are predominantly American in style.

If Intel integrates with GlobalFoundries, it could bring U.S. semiconductor manufacturing back together. Culturally and technologically, the two companies represent the combination best positioned to maximize integration benefits.

Moreover, Intel could learn from GlobalFoundries the know-how of running a foundry business, as well as the service mindset and customer engagement practices it has long lacked. By any measure, the two companies are a perfect match.

Of course, it will take time for a merger between Intel and GlobalFoundries to deliver results. However, if the integration proceeds smoothly, it could give them the chance to compete head-to-head with Asia’s TSMC and Samsung Electronics, while also helping TSMC avoid antitrust scrutiny. This would be a positive development for the industry as a whole.

I believe President Trump’s repeated claim that “Taiwan stole America’s semiconductor jobs” is wrong, and the statement is flawed on several levels. First, Taiwan has not taken away U.S. semiconductor jobs. The United States remains the global leader in the semiconductor industry, accounting for 43% of global semiconductor output, compared with Taiwan’s roughly 23%. In upstream areas such as IC design, intellectual property (IP), and EDA tools, the U.S. continues to hold an overwhelming advantage.

Second, Taiwan and the United States are allies, and their roles in the semiconductor industry are complementary rather than competitive. While TSMC’s rise has made life harder for Intel, it has created greater benefits for American companies such as Apple, NVIDIA, AMD, Broadcom, and Qualcomm, while also expanding semiconductor jobs in the U.S.

TSMC’s presence brings the greatest benefit to the United States. Moreover, only Taiwan has chosen to play the role of a “supporting alliance partner” in the industry; countries such as Japan, South Korea, and China have not taken this approach.

Taiwan’s path in developing semiconductors has been different from that of other countries. At the beginning, Taiwan focused on foundry services simply to survive, demonstrating a spirit not of defeating others but of succeeding together with customers. By excelling in wafer foundry manufacturing and building a strong alliance with U.S. industry, Taiwan has created the most ideal scenario for global industrial development.

The Premise and Experiment of Fair Competition

If President Trump still refuses to believe that the current situation is the best outcome, then why not run an experiment? Form a “Team USA” by combining Intel and GlobalFoundries, and pit it against Taiwan’s TSMC team and South Korea’s Samsung Electronics team in a fair final match, to see who survives in the end and who emerges as the winner by merit.

Of course, such a competition would need clear preconditions: it must adhere to the principles of fairness, justice, and transparency. There must be no cheating, and the Trump administration cannot act as both referee and rulemaker, altering the rules to its own advantage or secretly dealing Intel a winning hand. Doing so would violate the basic standards a contest is supposed to uphold.

If these principles are upheld, I believe TSMC will still emerge as the ultimate winner. If the Trump administration doubts this, it can run the experiment immediately—the results should become clear within three to five years.

In the longer term, however, the U.S. government’s stake in Intel could turn the company into a “national champion” and provide many favorable conditions, but it may also constrain Intel’s decision-making space and strategic autonomy.

This clause also means that if Intel seeks to spin off its foundry business, the decision is no longer solely the company’s own. It must take into account the Trump administration’s policies and views, as well as potential equity transaction conditions the government may set in the future. While Washington claims this is a “passive investment,” in reality it already wields significant influence.

For Intel, the funds provide temporary breathing room but may also constrain its strategic flexibility in the future. For U.S. semiconductor policy, the shift from subsidies to equity stakes strengthens control over domestic chip manufacturing, adding greater uncertainty to the industry’s future development.

More importantly, for the global foundry industry, if Intel truly spins off its foundry business and carries out meaningful reforms in the future, its situation would at least improve compared with today, and it could even reshape the market structure now dominated by TSMC and Samsung.

I believe TSMC has no fear of competition. If Intel performs better, the contest will be far more exciting; otherwise, as it stands now with TSMC pulling far ahead alone, the situation inevitably feels a bit dull.