Reports emerged this week confirming what many industry watchers have long suspected but been afraid to articulate: China has launched a semiconductor “Manhattan Project.” This massive, state-backed initiative has one existential goal – reverse engineering the ultra-complex lithography machines currently monopolized by the Dutch company ASML. By breaking the code of extreme ultraviolet (EUV) technology, Beijing aims to break the gate chains of US sanctions and achieve self-sufficiency in the production of chips needed to train pioneering artificial intelligence (AI) models.
This is not just large-scale industrial espionage; this is a geopolitical turning point. The race for AI supremacy is the most important fight of the 21st century, and the starting gun has just fired for the second time. According to exclusive reports from Reuters and Taiwan Newsin China has already built a working prototype in a high-security laboratory in Shenzhen that successfully generates EUV light. The consequences for U.S. leadership are profound, the scale of China's involvement is staggering, and the window in which the West can maintain its advantage is shrinking faster than policymakers realize.
A threat to American hegemony
For the past several years, the U.S. strategy for China's technological development has been based on armed interdependence. By controlling key bottlenecks in the semiconductor supply chain—particularly ASML's advanced chip design software and irreplaceable manufacturing tools—Washington has effectively limited China's ability to develop cutting-edge artificial intelligence. ASML's EUV machines are engineering marvels, using lasers to vaporize molten tin to create light waves capable of printing transistors as small as nanometers wide. These are the only tools capable of creating chips that power the most advanced models of generative artificial intelligence today.
If China's “Manhattan Project” succeeds in replicating or bypassing this technology, the main lever of American technological power will break. American semiconductor leadership is now defined not only by innovations like NVIDIA's, but also by the ability to deny adversaries access to those innovations. A self-sufficient China, armed with indigenous lithography capable of sub-7nm production, would immediately eliminate current export controls. This would unleash a flood of Chinese-developed artificial intelligence capabilities with direct military and economic applications, effectively ending the unipolar moment of technological dominance. As Asia Times notedthese efforts aim to kick the United States out of the supply chain entirely.
A tale of two scales: effort asymmetry
Comparing the West's efforts to revive chip production with China's new initiative reveals a surprising asymmetry. The United States celebrated the transition CHIPS and the Science Acta $52 billion package aimed at attracting manufacturers back to U.S. soil. While important in a Western legislative context, it is a market-based incentive program limited by political conflicts and corporate bureaucracy.
On the other hand, China's approach appears to be far superior to anything the West is currently undertaking. This is not a subsidy program; it is national mobilization on war grounds. According to reports, the project is supervised by, among others, Ding Xuexiang, a close ally of President Xi Jinpingand coordinated by Huawei. Beijing is implementing state capitalism with unlimited liability, recruiting former ASML engineers with huge bonuses and, in some cases, providing them false identities to avoid detection.
While the United States relies on private companies like Intel and Micron to make business decisions consistent with national security, China is directing state resources to solve a physical problem regardless of immediate return on investment. The Chinese government realizes that this is not about market share; it's about sovereignty. The scale of resources Beijing can devote to a single technological obstacle dwarfs the patchwork of incentives currently offered by the U.S. and its European allies.
A ticking clock for Western leadership
How much time do the United States and its allies have before China bypasses them? It's tempting to dismiss this effort by pointing to the enormous complexity of ASML machines, which took decades of global collaboration to perfect. ASML's CEO stated earlier this year that China would need it “many, many years” replicate this technology.
But underestimating China's technological speed is a historical mistake. Although the prototype currently has problems with optical precision and has not yet produced working chips, the schedule is tight. Sources close to the project indicate that the goal is to produce working chips by 2028, with a “realistic” target of 2030. This is potentially many years ahead of Western forecasts. If China devotes hundreds of billions of dollars and its best scientific minds to this problem, the West's ten-year advantage could be reduced to three to five years. Moreover, China may not need to replicate ASML machines perfectly; it just needs a “good enough” alternative that will allow it to train competitive AI models, even at higher costs and lower results. The danger zone for the West is not 2035; will start before 2030.
The American Imperative: Beyond Defense
To prevent a takeover, the United States must accept that defensive measures – sanctions and export controls – delay action, not a winning strategy. The current leaks in the sanctions system, as evidenced by China's ability to do so source components from secondary marketsshow that determined actors find solutions.
The US needs an offensive strategy that matches the urgency of China's mobilization. First, it requires dramatically increasing federal funding for research and development in next-generation semiconductor technologies, moving beyond silicon and focusing on areas such as advanced packaging and novel materials where the United States continues to lead. Second, the alliance with the Netherlands and Japan must be strengthened to ensure that the technology denial regime does not collapse under the pressure of the Chinese economy. Finally, the United States must win the war for talent by reforming immigration policy to ensure that the world's brightest engineers choose Silicon Valley over Shenzhen.
Summary
China's “Manhattan Project” semiconductor project is a clear signal that Beijing views technological dependence on the West as an unacceptable weakness and is willing to spend any amount to plug the gap. The United States can't win this race by simply trying to knock down another runner. It must move faster, invest deeper, and realize that the comfortable advantage it enjoyed in the silicon era is over. The age of artificial intelligence will be a centimeter by nanometer competition, and as these reports confirm, the race is much closer than we thought.
As President and Chief Analyst of the Enderle Group, Rob provides regional and global companies with guidance on creating a credible dialogue with the market, reaching customer needs, creating new business opportunities, anticipating technological changes, selecting suppliers and products, and practicing zero-dollar marketing. For over 20 years, Rob has worked for and with companies including Microsoft, HP, IBM, Dell, Toshiba, Gateway, Sony, USAA, Texas Instruments, AMD, Intel, Credit Suisse First Boston, ROLM and Siemens.
