The global race for semiconductor dominance is no longer a niche industrial topic. It sits at the intersection of economic power, national security and technological leadership. Over the next decade, who controls the design, production and supply of advanced chips will largely determine who leads in AI, 5G/6G, quantum computing, autonomous systems and modern warfare.
Behind every smartphone, every data center and every guided missile stand a handful of companies, a few strategic regions and a growing arsenal of export controls and industrial subsidies. The dependency is so strong that a single bottleneck at one site in Taiwan, the Netherlands or South Korea can paralyze entire sectors of the global economy.
So what exactly is at stake in this “chip supremacy” battle, who are the key players, and what are the concrete implications for businesses, states and citizens in the coming decade?
Why chips matter more than oil for the digital economy
Semiconductors are often described as the “new oil”. The analogy is imperfect, but the logic is similar: chips are a foundational input for almost every modern technology. Without them, value chains in automotive, telecoms, cloud computing, defence, finance and healthcare simply stop.
Three dynamics explain why their strategic value is rising so fast:
- Exploding demand from AI and cloud. Training a cutting-edge AI model like GPT-type systems can require tens of thousands of high-end GPUs running for weeks. Each GPU is built on some of the most advanced chips ever manufactured, with transistor densities at 5 nm, 3 nm and soon below.
- Electrification and digitisation of “old” industries. A modern electric vehicle can embed 1,500 to 3,000 chips. Industrial robots, power grids, medical imaging, logistics: everything gets sensors, connectivity and compute.
- Militarisation of computation. Advanced chips power drones, hypersonic missiles, missile defence systems, secure communications and cyber operations. They have become dual-use assets: civilian and military at the same time.
In other words, whoever controls advanced chips controls both the productivity curve of the global economy and the capability curve of modern armies.
Who controls what: a highly concentrated, fragile ecosystem
The chip industry is a global puzzle: no country, not even the US or China, controls the full stack. Power comes from dominating a few critical segments.
Broadly, the chain is composed of:
- Design (fabless companies like NVIDIA, Qualcomm, AMD, Apple, and design tools from US firms like Synopsys and Cadence)
- Manufacturing (foundries like TSMC in Taiwan, Samsung in South Korea, Intel in the US/Europe)
- Equipment and materials (ASML in the Netherlands for extreme ultraviolet (EUV) lithography, US and Japanese companies for other key tools and chemicals)
- Packaging and testing (strong presence in Southeast Asia: Malaysia, Vietnam, the Philippines)
Two concentration points are particularly critical:
- TSMC and Samsung produce practically all the most advanced logic chips (5 nm and below). According to various industry estimates, TSMC alone accounts for more than 50% of global foundry revenues and over 90% of the most advanced nodes.
- ASML holds a quasi-monopoly on EUV lithography machines, essential to produce the most advanced chips. Each machine costs over 150 million dollars, can take months to install, and is subject to strict export controls.
This extreme concentration makes the system efficient… and fragile. A geopolitical crisis around Taiwan, an export ban on Dutch or US equipment, or a cyberattack targeting a handful of suppliers could rapidly spread across entire sectors.
The US strategy: maintain the technological lead and weaponise interdependence
Since 2018, Washington has moved from a mostly laissez-faire stance to an openly interventionist approach on semiconductors. Two objectives dominate: stay ahead technologically and limit China’s access to the most advanced capabilities.
Key tools include:
- Export controls. The US has progressively restricted the sale to China of advanced chips, chip design software (EDA tools), manufacturing equipment, and even US-origin talent for certain highly sensitive projects. The October 2022 and 2023 packages target high-performance GPUs, advanced nodes and AI-related systems.
- Industrial policy. The CHIPS and Science Act (2022) mobilises around 52 billion dollars in subsidies and tax incentives to encourage semiconductor manufacturing and R&D on US soil, including major investments by TSMC, Samsung and Intel.
- Alliance building. Through informal frameworks like the “Chip 4” (US, Taiwan, South Korea, Japan) and coordination with the EU and the Netherlands, Washington seeks to align export controls and standards with its closest partners.
In practice, the US has a strong advantage in design (GPU, CPU, AI accelerators), software tools and key manufacturing equipment. Its strategy is to keep a 2–3 node lead over China and to turn this technological gap into geopolitical leverage: access to advanced chips becomes a bargaining chip in diplomatic and trade negotiations.
China’s response: massive investment and accelerated self-reliance
China is simultaneously the world’s largest importer and consumer of chips and one of the main challengers to US technological dominance. Semiconductors are at the heart of Beijing’s strategies “Made in China 2025” and “dual circulation”.
Faced with US controls, China is pursuing three parallel paths:
- Catch-up in domestic manufacturing. Firms like SMIC (Semiconductor Manufacturing International Corporation) and YMTC (Yangtze Memory) are trying to close the gap with TSMC and Samsung. In 2023, some reports suggested SMIC had achieved 7 nm production without EUV, but at higher cost and lower yield.
- Control of mature nodes and scale effects. Many industrial applications (automotive, IoT, industrial controls) still rely on 28 nm and above. China is investing heavily to dominate these segments, where the technological barrier is lower but volumes are high.
- System-level innovation. To compensate for lagging behind at the transistor level, Chinese players are focusing on system architectures, chiplet approaches and software optimisation, especially for AI models that can run efficiently on slightly less advanced hardware.
Estimates vary, but public announcements suggest that China has committed several hundred billion dollars (direct subsidies, tax breaks, state-backed funds) to its semiconductor ecosystem over the last decade. The key question is not the amount invested, but the ability to achieve genuine technological breakthroughs without access to the most advanced foreign tools and talents.
Europe, Japan, South Korea: between strategic autonomy and alliance with the US
Between the two giants, other players are trying to find their position, balancing industrial interests and security alliances.
Europe launched its own “EU Chips Act” with a target of reaching 20% of global chip production by 2030 (versus around 10% today). The bloc relies on:
- Existing strengths: ASML in lithography, Infineon and STMicroelectronics in power electronics, NXP in automotive chips.
- New mega-fabs: Intel in Germany, TSMC in Germany, possibly additional projects in France and Italy.
But Europe still lacks a true equivalent to NVIDIA, Qualcomm or AMD, and remains heavily dependent on Asian manufacturing for advanced logic chips.
Japan is trying to regain lost ground, with strong support to its domestic champions (Sony in image sensors, Renesas in automotive) and a new foundry project, Rapidus, aiming for 2 nm production in partnership with IBM and others. Its advantage: deep expertise in materials and equipment.
South Korea plays on both fronts: a key US ally, yet economically tied to China. Samsung and SK Hynix are global leaders in memory and advanced logic. Their investment decisions are influenced as much by subsidy packages as by export control risks.
From supply chains to security chains: the new risk matrix for states and companies
The semiconductor battle changes the nature of risk for both governments and companies. It is no longer enough to ask, “Can I secure my supply chain?” The question becomes, “How exposed am I to the geopolitical leverage of a small number of foreign actors?”
Some concrete vulnerabilities stand out:
- Single points of failure. A disruption in Taiwan (due to conflict, blockade or natural disaster) could halt advanced chip supplies for months. The impact would reach far beyond smartphones: cloud services, defence systems, industrial automation.
- Regulatory unpredictability. A new US export rule or a Chinese countermeasure can, overnight, render a product non-compliant or a supplier unusable. This affects not only defence-related firms but also civilian sectors relying on AI accelerators and networking gear.
- Data and cyber exposure. Chips are not just hardware; they embed firmware, microcode and secure enclaves. Who writes and updates that code? Under which jurisdiction? With what oversight? These questions matter for critical infrastructures and sovereign data.
For businesses, the risk is operational, financial and reputational. A car manufacturer facing chip shortages may halt production lines. A cloud provider cut off from high-end GPUs may fall behind on AI services. A telecom operator using equipment targeted by sanctions may need to rip and replace at massive cost.
The next 10 years: three major battlegrounds
Looking forward, three domains will likely structure the competition for chip supremacy and its implications for global security.
1. AI accelerators and data center chips
Today, NVIDIA dominates the AI training market, with GPUs like the H100 and its successors, while AMD, Intel and new players (Cerebras, Graphcore, Chinese actors like Huawei and Biren) try to carve out a space.
The stakes are high:
- Access to these chips determines who can train and deploy the most advanced AI models.
- Control over their export becomes a tool to slow down competitors’ AI capabilities, both economic and military.
- Cloud hyperscalers (AWS, Google, Microsoft, Alibaba, Tencent) are increasingly designing their own accelerators to reduce reliance on external suppliers.
We can reasonably expect a tri-polar competition: US-led ecosystems, China-centric stacks, and a more fragmented “rest of the world” trying to balance between both.
2. Edge, automotive and industrial chips
Not every strategic chip needs to be at 3 nm. For cars, grids, factories and cities, reliability, cost and security often matter more than ultimate performance.
This segment will see:
- Intense competition to localise production (to avoid a repeat of the 2020–2022 automotive chip crisis).
- Growing regulation on security-by-design for chips used in critical infrastructures (encryption, update mechanisms, resistance to tampering).
- A convergence between cyber regulations, safety standards and semiconductor industrial policies.
For Europe in particular, this is an area where strategic autonomy is more achievable than in bleeding-edge AI accelerators.
3. Quantum, photonics and post-silicon technologies
Beyond the current battle on 3 nm and below, another race is starting on alternative computing paradigms: quantum, neuromorphic, photonic chips. Their timelines are uncertain, but the implications for cryptography, optimisation and simulation are considerable.
Here again, the US, China and a few European and Asian countries are investing heavily. Even if commercial applications remain limited in the short term, breakthroughs in the next decade could shift the balance of power in fields like cryptanalysis and advanced military systems.
What this means for companies: from procurement to strategy
For many corporate leaders, semiconductors have long been a “technical” topic, left to engineering and procurement teams. This is no longer tenable. Chips are now a strategic variable that boards and executive committees must integrate into their risk management and long-term planning.
Concretely, several moves are worth considering:
- Map your semiconductor exposure. Which products depend on which types of chips, from which suppliers, in which countries, at which technology nodes? How concentrated is this exposure? Few firms can answer this precisely today.
- Scenario planning with geopolitics included. What happens to your business if Taiwan becomes inaccessible for 6–12 months? If US export controls tighten further? If China restricts exports of certain critical materials or components?
- Diversify not just suppliers, but geographies and nodes. Relying 100% on a single advanced node may be efficient but risky. In some cases, designing products that can run on slightly older, more widely available nodes may increase resilience.
- Engage proactively with policymakers. Subsidy programmes, export regimes and local content rules will shape the cost and feasibility of your semiconductor strategy. Staying passive is no longer an option.
In short, chips must be treated as a strategic resource, not just a line item in the bill of materials.
For states: between industrial ambition and strategic restraint
Governments face a delicate balance. On one hand, there is strong pressure to “onshore” or “friend-shore” semiconductor production, for jobs, resilience and security. On the other, full autarky is neither realistic nor efficient.
Some pragmatic principles are emerging:
- Focus on critical segments rather than everything. No country can be self-sufficient across the entire stack. Prioritising a few strategic nodes (e.g., advanced logic, power electronics, secure chips for defence) is more realistic.
- Build redundant capacity with allies. The goal is not to move everything home, but to ensure that no single geography can paralyse global supply. This implies coordinated investments and clear security frameworks between allied countries.
- Maintain open standards and interoperability. Fragmentation into incompatible technological blocs would be costly and could slow innovation. The challenge is to protect security without breaking the shared foundations that make global ecosystems productive.
For mid-sized powers, including many European countries, the most effective strategy often lies in focusing on niches (equipment, materials, design IP, trusted manufacturing) while leveraging alliances for the rest.
A decisive decade for technology and security
The battle for chip supremacy is not an abstract technological rivalry. It will shape who can deploy the most advanced AI models, who can secure their critical infrastructures, who can project military power, and who can absorb or impose economic shocks.
For businesses, ignoring this shift is a luxury that no longer exists. For policymakers, the question is less “Should we intervene?” than “How to intervene intelligently, without fuelling a spiral of protectionism and fragmentation?”
The next ten years will likely not bring full “semiconductor sovereignty” for anyone. But they will redraw the global map of dependencies, alliances and vulnerabilities. Those who take the time now to understand their position in this new landscape – and to adapt their strategies accordingly – will be better placed to navigate the turbulence ahead.