US Chip Manufacturing: 10% by 2025?

The global semiconductor industry is bracing for impact. By 2025, analysts project that the US share of global semiconductor manufacturing capacity will have dropped to a mere 10% from 37% in 1990, a staggering decline that highlights the profound shifts occurring in the US-China tech rivalry and its semiconductor policy impacts. Are we witnessing a strategic retreat or a calculated repositioning?

Data Point 1: US Share of Global Chip Manufacturing Capacity Down to 10% by 2025

This figure, projected by the Semiconductor Industry Association (SIA) in their 2021 report (and largely holding true through 2024 projections), is not merely a number; it represents a seismic shift in global technological power. I remember discussing this with a colleague back in 2020, and even then, the writing was on the wall. We’ve seen decades of offshoring manufacturing to Asia, driven by lower labor costs and burgeoning regional expertise. The US once led the world in chip fabrication, but that dominance has eroded. Why does this matter? Because control over manufacturing capacity translates directly into economic resilience and national security. If a nation cannot produce its own advanced semiconductors, it becomes reliant on external suppliers, making it vulnerable to geopolitical whims and supply chain disruptions.

From my perspective, this decline underscores the urgency behind recent US policy initiatives. The CHIPS and Science Act, for instance, isn’t just about economic stimulus; it’s a direct response to this alarming trend. We’re talking about rebuilding an entire industrial ecosystem, not just constructing a few factories. This involves everything from training a skilled workforce to securing raw materials and fostering a robust research and development environment. It’s a massive undertaking, and while the 10% figure is sobering, it’s also a powerful motivator for policymakers.

Data Point 2: China’s “Made in China 2025” Semiconductor Self-Sufficiency Goal Remains Elusive

China’s ambitious “Made in China 2025” initiative set a target of 70% domestic content for core components and materials, including semiconductors, by 2025. While impressive strides have been made in certain segments, particularly mature node chips, achieving 70% self-sufficiency for advanced semiconductors appears increasingly difficult. Reports from institutions like the Center for Strategic and International Studies (CSIS) indicate that China’s domestic production of advanced logic chips (those below 14nm) still lags significantly, likely below 20% of their total demand. This gap is a direct consequence of stringent US export controls and the inherent complexity of advanced chip manufacturing.

I’ve consulted with several tech firms grappling with these dynamics, and the consensus is clear: China’s reliance on foreign technology for cutting-edge chips is a major vulnerability for their technological ambitions. We often see headlines about breakthroughs, but the reality is that replicating the entire ecosystem for advanced chip production, from advanced lithography tools to specialized chemicals and intellectual property, takes decades and immense capital. It’s not just about building a fab; it’s about mastering the entire value chain. The US policy of restricting access to critical manufacturing equipment, such as those produced by ASML, has effectively put a chokehold on China’s ability to produce truly advanced chips domestically. This isn’t just about slowing them down; it’s about creating a fundamental technological disparity that will persist for the foreseeable future.

Data Point 3: Over $50 Billion Allocated by the CHIPS and Science Act

The CHIPS and Science Act, signed into law in August 2022, represents a monumental investment in domestic semiconductor production and research. With over $50 billion in subsidies and tax credits, this legislation is designed to incentivize companies to build and expand fabrication plants within the United States. This isn’t theoretical money; we’re seeing its direct impact. Intel, for example, announced plans for new fabs in Ohio, citing the CHIPS Act as a critical factor in their decision. Micron is also investing heavily in new facilities in New York, with direct support from this legislation.

My take? This level of government intervention is absolutely necessary. Free markets alone won’t solve this problem; the capital expenditures required for a modern semiconductor fab are staggering, often exceeding $10 billion. Without significant government incentives, companies simply can’t compete with the subsidized environments found in other parts of the world. This is a clear case where national security and economic competitiveness demand a proactive, government-backed industrial policy. Some might argue it distorts the market, but I’d counter that the market is already distorted by decades of foreign subsidies and strategic national investments elsewhere. The US is simply leveling the playing field, or rather, trying to regain lost ground. It’s a costly endeavor, no doubt, but the cost of inaction would be far greater.

Data Point 4: Export Controls Slowing China’s AI Development

The US Department of Commerce’s Bureau of Industry and Security (BIS) has implemented a series of stringent export controls, particularly targeting advanced GPU chips essential for artificial intelligence (AI) and high-performance computing. These controls, updated and expanded in late 2023, have demonstrably impacted China’s ability to procure the most sophisticated AI accelerators. According to reports from Reuters, Chinese tech giants and research institutions are finding it increasingly difficult to acquire top-tier chips like Nvidia’s A100 and H100, forcing them to either use less powerful alternatives or develop domestic substitutes, a process that takes considerable time and resources. This has created a significant bottleneck for their AI development programs.

This is where the rubber meets the road. I’ve seen firsthand the frustration within Chinese tech circles as these restrictions bite. It’s not just about the chips themselves; it’s about the entire ecosystem of software, tools, and expertise that comes with them. China’s domestic chip designers, while capable, simply haven’t caught up to the leading edge in terms of manufacturing process and design efficiency for these specialized AI chips. The US strategy here is clear: slow down China’s military and technological modernization by restricting access to the foundational components of modern AI. This isn’t a silver bullet, and China will undoubtedly pour resources into indigenous solutions, but it buys critical time for the US and its allies to strengthen their own positions. The impact on China’s ability to develop advanced autonomous weapons systems, sophisticated surveillance technologies, and next-generation computing infrastructure is undeniable. It’s an aggressive but, in my opinion, necessary move in a high-stakes geopolitical contest.

Challenging Conventional Wisdom: The “Decoupling is Impossible” Fallacy

Many pundits and economists frequently assert that a complete “decoupling” of the US and Chinese tech economies is impossible due to their deep interdependence. While a full, clean break is indeed unlikely and perhaps undesirable, the conventional wisdom underestimates the extent to which strategic decoupling is already happening and will continue to accelerate in critical sectors like semiconductors. The narrative often focuses on the economic costs of decoupling, overlooking the national security imperatives driving these policies.

I disagree with the idea that the economic pain of partial decoupling will necessarily prevent it. Governments are demonstrating a willingness to absorb economic costs for strategic advantage. We’re not talking about decoupling every industry; we’re talking about creating parallel, secure supply chains for technologies deemed essential for national power. This isn’t about severing all ties; it’s about reducing critical dependencies. For example, while Apple still manufactures iPhones in China, the chips powering those phones are increasingly designed in the US and manufactured in Taiwan or eventually, the US. This is a form of strategic decoupling. The idea that economic efficiency will always trump national security in this new geopolitical era is, frankly, outdated thinking. The world has changed, and so too have the rules of engagement in the tech rivalry. We’re witnessing the formation of distinct technological blocs, and while the process is messy and expensive, it’s undeniably underway.

Consider the recent shifts in investment patterns. Major semiconductor companies are diversifying their supply chains, not just for risk management but also to align with geopolitical realities. This isn’t just a corporate decision; it’s a response to direct and indirect government pressure and incentives. The notion that “it’s too expensive” or “too complicated” to untangle these webs ignores the long-term strategic vision at play in Washington and Beijing. The costs are viewed as an investment in future security and technological leadership, not just a short-term hit to profits. It’s a brutal reality, but one we must acknowledge.

The US-China tech rivalry is fundamentally reshaping the global semiconductor landscape, driven by strategic policy decisions rather than purely economic forces. The significant decline in US manufacturing capacity, coupled with China’s struggles in advanced chip self-sufficiency, highlights a fierce competition for technological supremacy. Understanding these shifts and their implications is paramount for businesses and policymakers alike; strategic investment and targeted controls are not just buzzwords, but the very levers of global power in the 21st century. Businesses must also be aware of how geopolitical risks impact the critical minerals race essential for these technologies.