Hope Is Not a Plan
The sustainability of advanced technologies, unique manufacturing capabilities, global access, and robust supply chains is currently dependent on ill-defined, reckless, and volatile political and economic strategies. Ignoring the reality of the situation and hoping things will eventually work out isn’t a good plan.
For decades, the world has relied on Taiwan Semiconductor (TSMC) to produce the most advanced chips, powering everything from smartphones to artificial intelligence. This dependence has created an unprecedented vulnerability: a single geopolitical flashpoint controls the lifeblood of the global digital economy.
While the United States and its allies express commitment to peace across the Taiwan Strait, relying on stability without practical preparation is dangerously shortsighted. Building resilience in the semiconductor supply chain requires tangible investments, diversification of sources, and coordinated industrial strategy. Without these steps, the U.S. risks supply shocks, price spikes, stalled innovation, and diminished technological leverage.
The Taiwan Dependency Problem
The numbers tell the story. TSMC manufactures over 90% of the world’s most advanced semiconductors, including chips smaller than 5 nanometers. These chips are essential for artificial intelligence, defense systems, high-performance computing, and next-generation consumer devices. No other company—Intel, Samsung, or otherwise—matches TSMC’s combination of technical sophistication, volume capacity, and yield efficiency.
This dominance creates extraordinary leverage. If Taiwan were disrupted by conflict, blockade, or even natural disaster, global technology supply chains would seize up almost instantly. The U.S., despite being the birthplace of semiconductor design, is structurally dependent on a single foreign company for chips that underpin its most advanced industries. Perhaps it is naïve vulnerability that has placed the US in this position, but it is unavoidable. Unfortunately, while there is loud rhetoric and saber-rattling, serious actions have been the equivalent of putting its head in the sand.
China, Technology Bans, and the Shifting Strategic Balance
The recent Chinese ban on companies purchasing Nvidia chips is a critical signal. For years, Washington assumed that export controls on AI hardware would restrain China’s progress. But Beijing has accelerated domestic semiconductor investment, poured billions into state-backed fabs, and mobilized its vast research community.
The ban is a much more potent signal that most people are ignoring. It suggests China believes it is far enough along in developing competitive chips—or at least confident enough in its trajectory—that it can push back.
This disrupts the strategic balance in global technology, marking a seismic shift. What was once a U.S. lever of control has become evidence of China’s growing independence. Washington now faces the reality that its dominance in chip design is eroding, and its dependence on Taiwan is increasingly untenable.
The U.S. Hits Its Limit
The questionable wisdom of industrial policy.
The CHIPS Act was an initial attempt to rebalance advanced technological manufacturing toward the US and shore up this vulnerability. The Act committed $50 billion to domestic semiconductor manufacturing.
Predictably, any time political involvement with meandering policy and focus, the process quickly became inefficient and misguided. Building advanced fabs in the United States is not just about pouring concrete—it requires supply chains for specialty chemicals, precision equipment, ultra-pure water, and most critically, skilled engineers and technicians.
It is not flipping a switch.
The United States faces a complex array of challenges, including labyrinthine requirements and approvals, labor issues, equipment and supply shortages, and, most critically, a severe shortage of skilled labor.
Meanwhile, TSMC’s planned facilities in Arizona have been delayed due to labor shortages and cost overruns. Intel’s roadmap remains uncertain, struggling with execution even as it attempts to regain technological leadership. Samsung is expanding in Texas, but scaling production at the most advanced nodes remains a daunting task.
These challenges underline the gap between aspiration and execution. Hope is not enough; execution must match rhetoric.
Now What?
Current strategies fall short in implementation. But what is a credible US strategy for semiconductor resilience? The unavoidable but difficult prescription must focus on several practical dimensions:
Talent Development
The bottleneck is not just capital—it’s human capability. Training semiconductor engineers, technicians, and materials scientists must become a national priority, supported by universities, vocational schools, and public-private partnerships.
Unfortunately, current restricted visa proposals, attacks on advanced educational institutions, and other unfortunate political machinations undermine perhaps the most critical component of success. The US is chasing away the best and brightest people and forcing the best institutions to cut back on research and development, as well as ongoing work. This is a recipe for failure.
Supply Chain Diversification
The U.S. and its allies must ensure multiple points of advanced chip production. This means accelerating fabs not only in the United States but also in Japan, South Korea, and parts of Europe. The goal is redundancy: no single chokepoint can jeopardize the system.
Allied Coordination
Harmonized policy across democracies. Japan is already co-investing with TSMC; the Netherlands plays a critical role with ASML’s lithography tools. Coordinated supply chain security agreements can spread costs and reduce duplication while strengthening resilience.
Investment Asymmetry
Simply catching up is not enough. The U.S. should invest aggressively in post-silicon technologies, including quantum computing, neuromorphic chips, optical interconnects, and 3D integration. Long-term leadership requires a combination of investing thoroughly in the most advanced technologies and leapfrogging those with disruptive and innovative ideas.
This requires significant capital, creativity, a willingness to fail, and a willingness to learn from those failures until great achievements that are hoped for on paper become a reality at the factory.
China
The more pressing question is whether Washington and Beijing can coordinate more effectively. Currently, antagonism prevails: the U.S. restricts exports, while China accelerates its self-sufficiency. This tit-for-tat risks creating two parallel, decoupled semiconductor ecosystems—inefficient, redundant, and ultimately destabilizing.
Is coordination possible? Perhaps only at the margins. Joint standards-setting, climate-focused chip applications, or shared investment in less sensitive technologies could help reduce the zero-sum framing. However, realistically, trust has eroded too far for deep cooperation in advanced semiconductor technologies.
The competition is here to stay.
Then What?
If cooperation proves impossible—and it likely will—then the U.S. must plan for a decoupled future. That means:
Nvidia must diversify manufacturing partners. While TSMC remains indispensable, Nvidia should strengthen its relationships with Samsung and explore co-development opportunities with emerging U.S. fabs. Nvidia may also need to segment product lines: leading-edge chips for domestic and allied markets, and slightly lagging versions for export to China or other restricted regions.
TSMC will remain a single point of failure for years. Washington must secure long-term guarantees for TSMC’s U.S. facilities while providing financial, logistical, and security support. It must also work closely with Taipei to ensure mutual trust, as TSMC’s interests cannot be taken for granted, particularly under the shadow of geopolitical pressure.
Beyond inefficient industrial policy and overly generous subsidies, the U.S. must consider stockpiling advanced chips, incentivizing the reshoring of critical supply chain components, and even exploring treaty-level frameworks with allies to ensure chip availability in times of crisis.
The Stakes
Semiconductors are not just another industry—they are the backbone of all advanced technologies, critical for economic growth, competitiveness, technological leadership, and military readiness. A disruption would cascade through global markets, from consumer electronics to cloud computing to defense systems.
Hoping Taiwan remains peaceful is not a plan. Hoping China’s rise in semiconductors falters is not a plan. Hope does not create fabs, train engineers, or align allied strategies.
What is required is a disciplined, sustained effort to diversify production, invest in talent and technology, and prepare for an era where the U.S. cannot take its semiconductor dominance for granted. This requires coordinated, nonpartisan efforts and public-private partnerships, along with an overall shared vision. It’s easy to be pessimistic about whether or not this can happen, but there is no other choice.
Resilience and Vulnerability
The challenges of advanced semiconductor technologies and manufacturing are among the most pressing and significant issues of this generation. The U.S. must acknowledge that a world dominated by a single supplier is unsustainable. It must invest not only in fabs but also in intellectual capital, allied coordination, and long-term technological leaps.
There is no guarantee of success. The rivalry with China will intensify, and Taiwan will remain a flashpoint. But inaction is the greater risk. Hope may provide comfort, but only strategy, investment, and execution will ensure resilience. Hope is not a plan.