In the 20th century, nations went to war over oil. In the 21st, the world’s next great resource conflict may be fought not over barrels of crude, but over nanometers of silicon. Just as oil powered the engines of industrial growth, semiconductors now fuel the intelligence of our digital age. From smartphones to supercomputers, and from autonomous vehicles to generative AI, chips are the invisible infrastructure on which the modern economy is built.
The explosion of artificial intelligence has accelerated this shift. Training large language models, running real-time financial simulations, or enabling advanced medical diagnostics requires computing power of an unprecedented scale. This demand is reshaping industries, inflating the valuations of chipmakers, and placing semiconductors at the heart of global power struggles.
But why are semiconductors being called the “new oil”? The answer lies not only in their economic importance, but also in their scarcity, their concentration in fragile supply chains, and the strategic leverage they grant to those who control them. Much like oil in the last century, chips are more than just a commodity: they are a geopolitical weapon and a driver of innovation.
In this article, we will explore the technological rise of AI chips, the geopolitical battleground they have created, and the economic and financial stakes of this new global race. The race for semiconductors is not just a story about technology—it is the story of who will lead the 21st century.
I. The Rise of AI Chips: From Silicon to Supercomputing Power
1. The Role of Semiconductors in AI Development
At the heart of artificial intelligence lies a simple truth: no algorithm, no matter how brilliant, can function without the right hardware. Just as oil powered engines and electricity powered factories, semiconductors power the intelligence of machines. They are the fundamental building blocks of modern computing, etched with billions of transistors smaller than a virus.
Traditionally, CPUs (Central Processing Units) were the workhorses of computation. But as artificial intelligence grew in complexity, CPUs proved too slow and too rigid. Enter GPUs (Graphics Processing Units), originally designed for rendering video games, which turned out to be perfectly suited for the parallel computations needed in deep learning. Today, when you ask ChatGPT a question or when Tesla’s Autopilot interprets road data, it is a GPU (most often from Nvidia) doing the heavy lifting.
Other specialized chips have since entered the scene:
- TPUs (Tensor Processing Units), developed by Google, are designed to accelerate machine learning tasks.
- FPGAs (Field-Programmable Gate Arrays) and ASICs (Application-Specific Integrated Circuits) are being tailored for narrow, high-performance AI applications.
- Startups are experimenting with chips optimized for neuromorphic computing, mimicking the human brain.
The demand for computing power has exploded alongside the rise of generative AI. Training a large language model can require tens of thousands of GPUs running for weeks, consuming massive amounts of electricity and capital. As Moore’s Law—the doubling of transistor density every two years—slows down, the industry is racing to build more efficient and powerful chips that can keep up with AI’s appetite.
2. Key Players in the AI Chip Race
The race for AI chips is both technological and financial, with a handful of companies leading the charge.
- Nvidia: Once a niche supplier for gamers, Nvidia now sits at the center of the AI boom. Its GPUs, combined with its proprietary CUDA software ecosystem, have become the gold standard for AI training. In 2024, Nvidia’s market capitalization briefly surpassed $3 trillion, making it one of the world’s most valuable companies.
- AMD: A long-time rival of Nvidia in the graphics space, AMD has made significant progress with its MI300 accelerators, which are increasingly being adopted in data centers. AMD’s openness and competitive pricing make it a key challenger.
- Intel: Once the undisputed king of CPUs, Intel has struggled to catch up in the AI race. However, it is investing heavily in Gaudi accelerators (through its Habana Labs acquisition) and betting on new architectures to regain relevance.
- Big Tech’s In-House Efforts:
- Google’s TPUs power much of its internal AI and cloud services.
- Amazon’s Trainium and Inferentia chips target AI workloads in AWS.
- Apple’s M1/M2/M3 chips demonstrate the power of vertical integration, optimizing AI for consumer devices.
- Emerging Players and Startups: Companies like Graphcore, Cerebras, and Tenstorrent are innovating with unique chip designs, such as wafer-scale engines that challenge the conventional limits of processing power.
The competition is fierce, not just for technological superiority but for market dominance. Owning the best AI chips doesn’t just mean higher profits—it means setting the pace of innovation for the entire global economy.
3. Why Chips Are Becoming the “New Oil”
The comparison between semiconductors and oil is not just rhetorical—it reflects deep economic and strategic parallels.
- Scarcity and Concentration
Just as oil reserves are unevenly distributed, advanced semiconductor manufacturing is concentrated in a handful of locations. Over 90% of the world’s most advanced chips are manufactured by Taiwan’s TSMC, creating vulnerabilities similar to those seen in oil chokepoints like the Strait of Hormuz. - Dependency Across Industries
Oil was the indispensable fuel of the 20th century; semiconductors are the indispensable logic of the 21st. Without them, industries as diverse as finance, healthcare, telecommunications, defense, and entertainment would grind to a halt. The push toward autonomous vehicles, smart cities, and digital economies only deepens this dependency. - Geopolitical Leverage
Countries that control chip production and design wield enormous strategic power. Just as OPEC once influenced global markets through oil supply, chipmakers and the nations that back them now hold sway over the future of AI, data, and security. - Catalyst for Growth
Oil powered industrial revolutions, but semiconductors are powering the digital revolution. They are the enablers of innovation, productivity, and new business models. Without them, there is no AI revolution, no Industry 4.0, no digital transformation.
Semiconductors are thus more than a technological resource—they are a strategic asset. The companies and countries that master AI chip production won’t just dominate markets; they will shape the balance of global power.
II. The Geopolitical Battleground of Semiconductors
1. The Fragile Supply Chain: Concentration and Risks
The semiconductor industry is one of the most globalized and complex supply chains in the world. A single advanced chip may require design software from California, photolithography machines from the Netherlands, rare materials from Japan, and final manufacturing in Taiwan or South Korea. This extreme interdependence makes semiconductors both a triumph of globalization and a glaring vulnerability.
At the center of this web sits Taiwan Semiconductor Manufacturing Company (TSMC). TSMC controls more than 90% of the world’s leading-edge chip production (below 5 nanometers). This dominance means that much of the global AI revolution is literally built on silicon etched in fabs near Hsinchu, Taiwan. The island’s geopolitical tensions with China thus translate directly into fears for global technological stability.
Other choke points exist:
- ASML, a Dutch company, is the world’s sole supplier of EUV (Extreme Ultraviolet) lithography machines, essential for producing cutting-edge chips. Without ASML, no country can manufacture the most advanced semiconductors.
- Japan and South Korea dominate critical materials like photoresists and silicon wafers.
- The U.S. and UK remain leaders in chip design (Nvidia, AMD, ARM).
This level of concentration is equivalent to energy markets depending on one or two countries for almost all oil production. Any disruption—whether from natural disasters, political instability, or armed conflict—could send shockwaves through the global economy.
2. The U.S.-China Tech War
Nowhere is the strategic importance of chips more visible than in the ongoing U.S.-China technological rivalry.
- Export Controls: Since 2022, Washington has imposed strict export bans on advanced chips and the tools needed to manufacture them, aiming to prevent Beijing from developing cutting-edge AI with potential military applications. Nvidia, for example, has been forced to develop less powerful versions of its GPUs to continue limited sales in China.
- China’s Push for Self-Sufficiency: In response, Beijing has doubled down on its “Made in China 2025” strategy. Companies like SMIC (Semiconductor Manufacturing International Corporation) and Huawei are investing billions in domestic production. Huawei’s recent success in producing a 7nm chip despite sanctions shows both the resilience and determination of China’s industry, even if it still lags behind TSMC.
- Economic Repercussions: These restrictions have global ripple effects. U.S. firms lose access to one of their largest markets, while Chinese companies scramble for alternatives. Multinationals with supply chains spanning both countries face growing uncertainty.
This tech war is not simply about profits—it is about strategic control over the future of artificial intelligence, defense, and data sovereignty. Just as oil once shaped alliances and conflicts, chips now define the new balance of power.
(For deeper analysis, see the Council on Foreign Relations’ coverage of the U.S.-China technology rivalry.)
3. Alliances and Fragmentation of the Global Market
Semiconductors are reshaping not only rivalries but also alliances. Governments worldwide are treating chips as a national priority, leading to a surge in industrial policies and subsidies.
- United States (CHIPS and Science Act): Passed in 2022, this $52 billion initiative aims to boost domestic chip production and reduce reliance on East Asia. Intel, TSMC, and Samsung are all building massive fabs in Arizona and Texas, fueled by federal support.
- European Union (EU Chips Act): Europe, which produces less than 10% of the world’s chips, aims to double its share by 2030. Partnerships with Intel and TSMC are central to this strategy.
- Japan and South Korea: Both nations are investing heavily to maintain their edge in materials and manufacturing. Japan, in particular, is revitalizing its semiconductor sector through state-backed projects like Rapidus.
- India: Aspiring to become the “world’s next semiconductor hub,” India is offering incentives to attract manufacturers, though it still faces major challenges in infrastructure and expertise.
This race has another consequence: fragmentation of the global market. If supply chains split into competing blocs—one led by the U.S. and allies, another by China—the result could be higher costs, inefficiencies, and reduced innovation. Just as the oil market was shaped by geopolitical blocs, semiconductors may follow the same path, with countries choosing their “chip alliances.”
Europe has launched the EU Chips Act to reduce dependency, yet its startups still face structural hurdles in scaling globally—echoing the broader challenges we highlighted in Why Do European Unicorns Struggle to Compete with Silicon Valley?
For investors, corporations, and policymakers, this fragmentation is both a risk and an opportunity. Governments are betting billions to ensure they are not left behind in what could be the defining technological contest of the century.
III. Economic and Financial Stakes of the AI Chip Race
1. Investment Boom: From Venture Capital to Mega Fabs
The race for semiconductors is capital-intensive to an almost unimaginable degree. Building a single advanced fabrication plant (“fab”) can cost between $15 and $25 billion, and the most sophisticated facilities—like those of TSMC or Samsung—require years to complete and constant reinvestment. Unlike oil wells, which can produce for decades, fabs become obsolete within a handful of years as technology progresses.
This enormous capital expenditure (capex) means that only a small number of players can compete at the frontier. Governments are stepping in with subsidies to keep fabs onshore:
- The U.S. CHIPS and Science Act provides $52 billion in incentives.
- The EU Chips Act allocates over €40 billion to boost European production.
- China, meanwhile, has created massive state-backed funds (the so-called “Big Fund”) worth over $150 billion to drive domestic chip development.
Private capital is following. Venture capital firms are pouring money into startups like Cerebras and Graphcore, betting on breakthrough architectures that could challenge Nvidia. Meanwhile, Wall Street and institutional investors are riding the semiconductor wave: Nvidia, TSMC, and AMD have all seen their valuations soar, creating both opportunities and fears of overheating.
One striking illustration: in mid-2024, Nvidia’s market cap briefly surpassed that of Saudi Aramco, the world’s largest oil company. This symbolic moment underscored how financial markets now view semiconductors not just as a cyclical industry, but as a core pillar of the future economy.
2. Chips and the Future of Global Industries
The economic importance of chips goes far beyond the semiconductor sector itself. Access to advanced AI chips is becoming a determinant of competitiveness across virtually all industries.
- Finance: High-frequency trading, risk modeling, and algorithmic portfolio management rely on ever faster, more powerful computing. Investment banks and hedge funds now see access to top-tier GPUs as a strategic advantage.
- Healthcare: AI-driven diagnostics, drug discovery, and genomics require enormous computational resources. Without advanced chips, breakthroughs in personalized medicine would stall.
- Automotive: Autonomous driving systems from Tesla, Waymo, and others depend on chips capable of processing terabytes of sensor data in real time. Semiconductors are becoming as vital to cars as engines once were.
- Defense and Security: Military applications of AI—from drone swarms to cyber defense—make chips a cornerstone of national security. This is why governments treat semiconductor access with the same seriousness they once reserved for oil reserves.
The ripple effects extend even further. Small and medium-sized enterprises (SMEs) in manufacturing, retail, and logistics increasingly rely on AI tools hosted in the cloud. The availability and price of chips directly influence the cost of adopting these technologies, and by extension, global productivity growth.
(For a broader perspective, see McKinsey Global Institute’s analysis on AI’s economic potential, which estimates AI could add up to $4.4 trillion annually to the global economy.)
3. Risks, Bubbles, and the Future Outlook
Where there is hype and extraordinary growth, there are also risks. The AI chip boom faces several financial, structural, and environmental challenges.
- Market Overheating
The meteoric rise of semiconductor valuations has led some analysts to warn of a potential bubble. If AI adoption slows, or if competition erodes margins, today’s astronomical market caps could correct sharply. The parallel with past tech bubbles looms large. - Overinvestment in Capacity
Governments around the world are simultaneously subsidizing new fabs. If too many come online at once, the result could be overcapacity, pushing down prices and profits. The semiconductor industry has historically been cyclical, and there is no guarantee that AI demand alone will smooth out those cycles. - Environmental Costs
Chip production is one of the most resource-intensive industrial processes on Earth. A single fab can consume millions of liters of ultrapure water per day, vast amounts of electricity, and rare earth minerals whose extraction has environmental and social costs. In an era of climate commitments, the sustainability of the chip boom is an open question. - The Technology Horizon
Beyond current AI accelerators, new paradigms are on the horizon. Quantum computing could one day render today’s GPUs obsolete for certain tasks. Neuromorphic chips, designed to mimic the brain, promise efficiency breakthroughs. Betting heavily on current architectures may prove risky if disruptive alternatives emerge faster than expected. - Geopolitical Shocks
As discussed in Part II, supply chain concentration makes the industry vulnerable. A conflict in the Taiwan Strait, for example, would not only disrupt global trade but also trigger financial turmoil on an unprecedented scale.
(For detailed industry trends and forecasts, see reports from the Semiconductor Industry Association.)
Conclusion
In the 20th century, oil fueled industrial might, geopolitical power, and financial growth. In the 21st, semiconductors have assumed that role. They are the indispensable infrastructure of the digital economy, powering everything from generative AI to autonomous vehicles, from medical diagnostics to financial trading. Without chips, the world’s most transformative innovations would remain ideas on paper.
As we have seen, the rise of AI chips is not only a technological evolution but also a concentration of power in the hands of a few companies and nations. Nvidia’s dominance in GPUs, TSMC’s near-monopoly on advanced manufacturing, and Big Tech’s in-house chip strategies all highlight how central semiconductors have become to innovation itself.
This concentration explains why semiconductors are no longer just a business concern—they are a geopolitical battleground. The U.S.-China tech war, global subsidy races, and the vulnerability of supply chains centered on Taiwan underscore how chips are shaping alliances and rivalries. Much like oil once dictated military strategy and trade policy, semiconductors now influence national security doctrines and global diplomacy.
The economic stakes are just as high. Trillions in market capitalization, billions in government subsidies, and the competitiveness of entire industries hinge on access to the most advanced chips. Yet this race also brings risks: potential bubbles, overcapacity, and significant environmental costs. The sustainability of this chip-fueled growth is an open question.
Ultimately, semiconductors are not simply the “new oil”—they may be even more critical. Oil powered machines; chips power intelligence. Whoever controls them will not only command markets but also shape the trajectory of global progress. The question is not whether semiconductors will define the 21st century—it is who will control this new resource, and at what cost to the world economy and society.
