The tiny transitor changed everything. When scientists at Bell Labs first demonstrated it in December 1947, nobody imagined that small silicon devices would one day control world power itself.
Today, nations compete fiercely over who designs, manufactures, and controls different segments of the semiconductor industry. Taiwan dominates advanced manufacturing, the United States leads in chip design and software, South Korea controls memory production, Europe holds critical lithography technologies, and China is racing to close the gap. This fragmentation has turned microchips into a central axis of geopolitical competition.
Control over advanced semiconductors now influences diplomatic leverage, alliance structures, and economic resilience. Export controls, industrial policy, and supply-chain security have become as strategically significant as troop deployments or weapons systems. The stakes are enormous. Whoever controls advanced chip manufacturing controls artificial intelligence, military superiority, and economic dominance.
Microchips: From transistors to tiny titans
The first transistor was crude; it looked like something held together with wire and hope. Three scientists earned a Nobel Prize for creating it: William Shockley, John Bardeen, and Walter Brattain.
Over decades, transistors became smaller. Engineers figured out how to stack thousands, then millions, then billions of them onto single chips. In 1965, Gordon Moore predicted that the number of transistors on a chip would roughly double every one to two years, a trend that—despite slowing—has guided the semiconductor industry for decades.
Today’s most advanced processors integrate tens of billions of transistors fabricated at scales of just a few nanometers, approaching the size of biological viruses and pushing the limits of physics and materials science. Each generation compounds available computing power, enabling tasks once considered impossible.
Chips now underpin nearly every critical system in modern society. Smartphones, automobiles, hospitals, power grids, financial markets, and military platforms all depend on semiconductor design and manufacturing. Recent global chip shortages have shown how disruptions can halt car production, delay medical equipment, and expose the fragility of highly concentrated supply chains. In the twenty-first century, semiconductor innovation is no longer just a technological concern—it is a foundational pillar of economic stability and national security.
Netherlands: ASML’s impossible grip on advanced chip manufacturing
Making advanced chips requires the most precise machinery humans have ever built. The extreme ultraviolet (EUV) lithography systems produced by Dutch company ASML are among the most complex machines ever assembled. Lithography is the process used to project intricate circuit patterns onto silicon wafers, defining where billions of transistors will be formed; without it, modern microchips cannot be manufactured.
Each EUV system integrates over 100,000 components sourced from thousands of specialized suppliers across multiple countries, reflecting a supply chain so fragmented that no single nation can replicate it independently.
An individual EUV machine costs between US$350 million and US$400 million, weighs close to 180 tons, and must be shipped in hundreds of crates before being assembled inside ultra-clean fabrication plants. Developing EUV lithography required more than three decades of sustained research and tens of billions of dollars in investment.
ASML has no real competition. The company controls over 90 percent of the world’s EUV lithography market, making it the gateway to advanced chip manufacturing.
A critical dependency lies in Germany. Carl Zeiss SMT supplies the ultra-precision mirrors used inside EUV machines. These mirrors must be manufactured with atomic-scale accuracy, deviating by no more than a fraction of a nanometer. At present, no other company is capable of producing optical systems at this level, making Zeiss an irreplaceable pillar of the advanced semiconductor supply chain.
Nvidia’s software lock-in strategy
American company Nvidia dominates data center chip design for artificial intelligence, controlling roughly 80% of the market. But Nvidia’s real power comes from software, not hardware.
At the center of this ecosystem is CUDA, Nvidia’s proprietary parallel computing platform and programming model. CUDA has become the de facto standard for AI development, deeply embedded in the tools, libraries, and workflows used by researchers and companies across the industry. While alternatives exist, none match CUDA’s maturity, performance optimization, or breadth of support.
Nvidia continually expands CUDA’s capabilities, adding new libraries, optimizations, and features that further entrench its position. Each iteration increases switching costs and extends customer dependence, reinforcing long-term lock-in rather than tying users to a fixed timeline.
This software-driven strategy gives Nvidia an advantage that goes beyond chip ownership. Control over the software layer allows the company to shape how AI is built, optimized, and deployed—making its influence more resilient than any single generation of hardware and harder for competitors to dislodge.
Taiwan’s semiconductor stranglehold
Taiwan produces more than 90 percent of the world’s most advanced semiconductors through a single company: TSMC. This concentration has turned an island of roughly 23 million people into a central pillar of the global digital economy, supplying technology on which every major industrial power depends.
Companies such as Apple and Nvidia rely almost entirely on TSMC to manufacture their most advanced chips. Even Intel, long a vertically integrated manufacturer, now depends on TSMC for parts of its cutting-edge production. Thousands of other firms across sectors ranging from smartphones to data centers are similarly tied to TSMC’s fabrication plants. In practice, the company operates like a nervous system for global technology.
If Taiwan’s advanced chip production were disrupted, the effects would cascade through the global economy. Shortages would first hit industries dependent on high-performance processors, including artificial intelligence, advanced weapons development, and data infrastructure. Over time, the disruption would spread to automobiles, medical equipment, and critical industrial systems, exposing how deeply modern economies depend on a narrow and geographically concentrated supply chain.
China’s secret Manhattan Project
At the same time, Chinese research teams are racing to develop indigenous advanced chip-manufacturing technology, with a particular focus on extreme ultraviolet (EUV) lithography. Multiple reports indicate that these efforts have recruited former ASML scientists, most notably Lin Nan, a senior researcher who previously served as a light-source competence owner within ASML’s EUV program.
Lin left the Netherlands in 2021 and returned to China, where he joined the Shanghai Institute of Optics and Fine Mechanics under the Chinese Academy of Sciences. There, he founded an advanced photolithography research group focused on EUV light-source development. Within eighteen months of his return, Lin’s team filed eight patent applications related to laser-produced plasma (LPP) EUV light sources for lithography and metrology, closely aligning with timelines identified in open-source reporting.
The group has also published research demonstrating solid-state laser–driven EUV light generation with conversion efficiencies reaching 3.42 percent in 2025—an important technical milestone, though still below the performance levels achieved in ASML’s commercial systems.
Chinese sources further indicate that research teams involving former ASML engineers completed a working EUV lithography prototype in early 2025 at a secure facility in Shenzhen. The system is reported to generate 13.5-nanometer EUV light, marking a significant step forward, though it is not yet capable of producing advanced semiconductor chips. Analysts describe the effort as a combination of reverse engineering, long-term state funding, and aggressive talent recruitment, with incentives reportedly reaching millions of yuan.
Despite these advances, experts caution that replicating ASML’s full EUV lithography capability remains a formidable challenge. ASML executives and independent analysts alike emphasize that while generating EUV light is achievable, integrating it into a complete, high-yield manufacturing system requires extreme precision across optics, materials, software, and system reliability. Even optimistic assessments suggest that achieving functional equivalence with ASML’s Twinscan machines would take many years, with realistic timelines extending toward the end of the decade.
Taiwan’s paradox and future tensions
Taiwan’s semiconductor dominance, led by TSMC, creates a dual strategic effect. Its control over the world’s most advanced chip production acts as a deterrent against Chinese invasion—the so-called “silicon shield”—by raising the global economic and security costs of conflict. At the same time, this concentration makes Taiwan a critical strategic prize as semiconductors underpin artificial intelligence, military systems, and economic power.
China’s drive for semiconductor self-sufficiency complicates this balance. US export controls on Huawei, significantly expanded in 2022 and tightened further with AI-focused restrictions in 2025, are intended to slow China’s access to advanced chips and manufacturing equipment such as ASML’s EUV systems. In parallel, US industrial policy—most notably efforts to expand advanced semiconductor manufacturing on American soil through companies such as Intel and TSMC—aims to reduce the economic shock of a potential war or annexation of Taiwan, and to prevent the United States from becoming dependent on China for advanced semiconductors in such a scenario.
Taiwan thus faces a strategic dilemma. Diversifying production through TSMC fabs in the United States, Japan, and Europe improves supply-chain resilience and aligns Taiwan with allied industrial strategies, but also risks eroding its unique strategic value. Meanwhile, US restrictions have slowed—but not stopped—China’s progress, as domestic efforts continue through state-backed research and talent recruitment. Most analysts expect China to narrow the technology gap toward the end of the decade.
Geopolitical standoff: the race accelerates
By 2024–2026, the semiconductor conflict moved from abstraction to execution. In November 2024, Intel finalized a US$7.86 billion award under the CHIPS and Science Act, anchoring a wave of fabrication projects across Arizona, New Mexico, Ohio, and Oregon. Combined with investment tax credits, Intel’s total US spending now exceeds US$100 billion. The objective is not technological dominance alone, but strategic insulation: reducing the economic and military shock the United States would face in the event of a war or annexation of Taiwan.
Export controls followed the same logic. In May 2025, Washington moved to block Huawei’s Ascend AI chips globally, a step reinforced weeks later when Taiwan placed Huawei and SMIC on its own trade blacklist. These actions hardened a bifurcated technology system. China accelerates indigenous development, while the United States and its allies tighten trusted supply chains. Beijing’s response—export controls on silver by late 2025 and expanded restrictions on rare earths—made clear that the contest now extends beyond chips to the raw materials essential for semiconductors and modern weapons.
China’s political signaling has advanced in parallel. In January 2026, President Xi Jinping declared Taiwan reunification “unstoppable,” while military exercises simulating a blockade underscored Beijing’s willingness to apply pressure short of invasion. Official timelines pointing to 2049 suggest not a fixed deadline, but a belief that technological dependence can be reshaped well before then.
What emerges is a strategic paradox. The more Washington succeeds in reducing its reliance on Taiwan-centered semiconductor production, the less economically catastrophic a conflict becomes for the West—and the more maneuvering room that creates for escalation. At the same time, China’s push toward self-sufficiency erodes Taiwan’s long-term economic leverage while potentially increasing short-term incentives to act before those capabilities fully mature.
The semiconductor war is no longer metaphorical. It is already being fought through factories, export controls, materials embargoes, and industrial policy. Whether this contest ultimately stabilizes the Taiwan Strait or accelerates confrontation will depend less on who masters the next nanometer—and more on how much dependence the global system dismantles before deterrence gives way to calculation.