← Companies | semiconductors

ASML

ASML: The Only Company That Can Print the World's Most Advanced Chips

| semiconductors
↓ .md Take this into your AI — the full analysis + graph as markdown, ready to paste into ChatGPT, Claude, Gemini or any AI.

Based on 149 related nodes across 23 research explorations in the semiconductors sector.

What ASML Actually Does

Imagine you want to print text so small that a thousand lines of it would fit across a single human hair. That is roughly what chip manufacturers do — they “print” billions of microscopic transistors onto silicon wafers. To do this at the smallest sizes, you need a very specific kind of machine that uses extreme ultraviolet light, almost like a camera that takes impossibly precise photographs onto silicon.

ASML, a Dutch company headquartered in Eindhoven, is the only company in the world that makes these machines. Not the best company — the only company. Every advanced chip in every AI server, every smartphone, every modern computer either came from an ASML machine or from a less-advanced chip that needed one to eventually exist.

That is not an accident or a temporary lead. It took 30 years and roughly $9 billion in research and development to build, and the machines themselves require parts that only one other company in the world can make (a German optics firm called Zeiss). No competitor has cracked this in three decades of trying.

The Lock-In Goes Deeper Than You Think

Here is the non-obvious part: ASML is not just a supplier. It has become a piece of geopolitical infrastructure.

The United States, Japan, and the Netherlands have quietly coordinated to use ASML — and ASML alone — as the primary tool to prevent China from building the world’s most advanced chips. The US can tell ASML which countries it is allowed to sell to, because ASML’s machines contain American technology. The Dutch government issues the actual export licenses. Japan controls the specialized chemicals chips need. Together, these three countries form an informal alliance, and ASML is the Dutch pillar holding up that structure.

This means ASML is simultaneously a private company and a strategic weapon. It cannot sell to whoever it wants. But it also receives political protection that no ordinary company would get — because if ASML were disrupted, the entire strategy of keeping advanced chips out of adversarial hands would collapse overnight.

Strengths

The monopoly compounds over time. ASML is not standing still. Its newest machines — called High-NA EUV — can print features even smaller than the current generation, enabling chips at scales (sub-2 nanometers, roughly the size of a few atoms across) that no other manufacturing method can reach. Because ASML is already ahead, each new generation it releases widens the gap faster than any competitor can close it. China’s chip manufacturers currently need nine times as many manufacturing steps per chip layer as ASML’s customers do. That structural cost disadvantage cannot be engineered away — it compounds.

A remote kill switch. This is one of the most structurally unusual facts in the brief: ASML machines in Taiwan — where the world’s most advanced chips are made by TSMC — contain a remote disable capability. If Taiwan were ever invaded or seized, ASML could render those machines inoperable from the Netherlands. This is not a rumor; it is a documented feature. It means that even the worst geopolitical scenario does not result in an adversary gaining productive use of the world’s most advanced chip-making equipment.

Every allied reshoring program depends on ASML. The United States is spending hundreds of billions of dollars to build advanced chip factories on American soil. Japan is building a new national chip company from scratch. The EU has its own chip sovereignty ambitions. Every single one of these programs requires ASML equipment to build the most advanced chips. Demand is not optional — it is baked into sovereign industrial policy across multiple decades.

Vulnerabilities

China is currently about 25-35% of ASML’s revenue. ASML still sells older-generation machines (called DUV) to Chinese chip manufacturers, and it services machines it sold years ago. If US legislators pass a pending law called the MATCH Act — which would ban both new sales and ongoing service to Chinese chip makers — ASML would lose a large chunk of its income quickly. The allied reshoring programs will eventually replace this revenue, but there is a timing gap: factory construction takes years, and the revenue loss would be immediate.

The monopoly has a prototype-level challenger for the first time. In December 2025, a Chinese company called SiCarrier — linked to Huawei — demonstrated a working EUV prototype, apparently built with the help of former ASML engineers. It is not a commercial product. It cannot yet make chips at scale. But it is the first time in the company’s history that anyone outside ASML has built a machine that generates extreme ultraviolet light for chip manufacturing using a different technical approach. The gap between “we built a working prototype” and “we are making chips competitively” is enormous — but it is no longer infinite.

ASML does not fully control its own fate. The US government has effective veto power over ASML’s customer decisions. If the US administration wanted to trade chip access to China for tariff concessions or a trade deal, ASML would have little ability to resist. There are already signs of tension between the legislators pushing stricter controls and a presidential administration that sometimes sees chip access as a bargaining chip in trade negotiations.

One upstream supplier is the entire supply chain. ASML’s EUV machines require precision optical components made only by Zeiss, in Germany. ASML owns about a quarter of Zeiss, but not a controlling stake. If anything disrupted Zeiss production — a fire, a political dispute, a change in ownership — ASML’s ability to ship machines would stop. This dependency is not a hypothetical risk; it is a structural single point of failure.

Bull Case: Why ASML’s Position Gets Stronger

The strongest argument for ASML is that every major trend reinforces its position rather than eroding it.

AI requires the most advanced chips, which require ASML machines. Allied governments are legally committed to building fabs that require ASML equipment. The export controls that restrict ASML’s China sales also prevent China from leapfrogging ASML’s monopoly by buying finished chips instead of building its own. And each new generation of ASML technology makes the gap wider — so the company is not defending a static position, it is advancing a moving one.

China’s own behavior confirms the controls are working. When a country is not worried about a technology gap, it does not launch emergency national programs to close it. China is pouring money into domestic chip development at a scale that only makes sense if existing controls are genuinely restricting its access. This is evidence that ASML’s chokepoint function is real, not theoretical.

The kill switch changes the strategic calculus around Taiwan. The scenario most threatening to the global chip supply — conflict in the Taiwan Strait — is significantly less threatening to ASML’s long-term position if TSMC’s machines can be remotely disabled. Physical disruption does not translate into strategic capability transfer.

Bear Case: Why ASML’s Position Could Erode

The strongest argument against ASML is that it faces pressure from two directions simultaneously — and its own best defensive tool accelerates the threat it is trying to contain.

The paradox: every time ASML (or the US government) denies service or spare parts to Chinese chip makers, it increases the urgency and scale of Chinese investment in building a domestic alternative. The more aggressively the alliance uses ASML as a weapon, the faster it motivates the one development that would make ASML’s monopoly obsolete. The SiCarrier prototype was built with former ASML engineers, partly because ASML itself trained a generation of Chinese employees before export controls tightened. The talent transfer already happened.

The coalition holding ASML in place is more fragile than it looks. The Netherlands’ willingness to enforce export controls depends on US political continuity and on the Dutch government absorbing whatever economic retaliation China chooses to apply to Dutch agricultural and industrial exports. A US administration willing to trade chip access for a favorable trade deal could crack the entire architecture. There are documented tensions between legislative push for stricter controls and executive branch commercial instincts.

And on a longer horizon — 10 to 15 years — quantum computing represents a path to advanced computation that does not require ASML’s machines at all. Quantum hardware uses older, cheaper fabrication processes. If quantum computing achieves practical use in AI or cryptography, the significance of leading-edge classical chip manufacturing diminishes. This is not an imminent threat, but it is a structural one.

Leverage Points Worth Understanding

ASML has three structural levers that most companies do not have at all.

First, it controls who gets the most advanced machines — and there are only 6 to 8 of the newest generation available each year. Deciding whether Intel, TSMC, Samsung, or a new Japanese fab gets one of those machines shapes which companies can lead in advanced chip manufacturing for the next five years.

Second, it can degrade the performance of Chinese chip makers without banning anything new, simply by gradually restricting spare parts, software updates, and on-site maintenance for machines already installed in China. This is a slow-acting lever, but it works, because the machines degrade in yield and throughput without ongoing support.

Third, the kill switch is most valuable when unused. The mere credibility that ASML can disable machines in a crisis reduces the incentive for anyone to seize those machines by force. The deterrent works precisely because everyone knows it exists and believes it would be used.

Bottom Line

ASML is the single most concentrated chokepoint in the global technology supply chain. Its monopoly on advanced chip-printing machines is genuine, physically embedded in hardware, and backed by a tripartite geopolitical alliance that has converted a commercial technology advantage into a strategic architecture for controlling who gets to manufacture the most advanced chips.

The monopoly is durable but not permanent. For the first time, a Chinese competitor has demonstrated a working prototype using a different technical approach. The talent that built ASML’s original machines has partially transferred. The political coalition maintaining the export controls faces pressure from both commercial interests and Chinese retaliation incentives.

The near-term financial risk is concrete: pending US legislation could cut 25 to 35 percent of ASML’s revenue by banning sales and servicing to China’s chip makers. The long-term technology risk is structural: a Chinese company that achieves even older-generation EUV capability eliminates the most urgent part of the competitive gap, even if it cannot match ASML’s frontier machines.

The most non-obvious finding: ASML’s position is strongest when its coercive tools are used with restraint. Aggressive service denial accelerates Chinese domestic development. Moderate, calibrated restriction maximizes yield degradation at Chinese fabs while buying time for allied reshoring to generate the demand that replaces China revenue. The company’s leverage is highest when its power is visible but not fully exercised.