Chip Supply Squeeze Keeps ASML at Center of AI Capacity Race

Chip supply is staying tight as ASML warns that the wave of artificial-intelligence spending now reaching logic, memory, satellites, and robotics is outrunning how fast the semiconductor industry can add capacity. The message matters because ASML sits at the center of the tool chain that its biggest customers need to expand.

In a Reuters interview published on May 20, ASML chief executive Christophe Fouquet said the market would remain supply-limited for quite a while, even as the company pushes more output into the field and works on new products. That warning lines up with ASML’s April 15 first-quarter results and investor call, where the company said chip demand was outpacing supply, customers were accelerating capital spending, and 2026 revenue guidance had been raised to a range of 36 billion euros to 40 billion euros.

Why Chip Supply Is Staying Tight

The new ASML comments matter because they shift the semiconductor conversation away from a short-lived AI surge and toward a longer phase of industrial constraint. For investors and operators, the issue is no longer whether demand is strong, but whether enough tools, cleanroom space, power, and specialized manufacturing capacity can be added in time.

ASML’s own numbers show why the market is taking that view seriously. The company reported first-quarter net sales of 8.8 billion euros and net income of 2.8 billion euros, then said the industry’s growth outlook was solidifying as AI-related infrastructure spending pushed customers to bring forward expansion plans for 2026 and beyond.

AI Demand Is Raising the Chip Supply Baseline

Fouquet told Reuters that AI demand is coming so strongly that the market will remain supply-limited for quite a while. He linked that pressure not only to conventional data-center buildouts, but also to projects such as Elon Musk’s proposed TeraFab chip complex and the continuing expansion of Starlink, both of which point to a wider range of compute-hungry end markets than investors were discussing a year ago.

That broader demand base matters because it changes the shape of the cycle. If AI were concentrated only in a few cloud companies, suppliers could hope for a more familiar capex peak followed by digestion. Instead, ASML is describing a market where the same underlying chip constraint is being reinforced by hyperscalers, foundries, memory producers, communications systems, robotics, and adjacent industrial platforms at the same time.

The result is that chip supply pressure now looks structural rather than episodic. As long as more sectors compete for leading-edge logic and memory capacity, equipment makers like ASML remain one of the clearest ways to read whether the industry can keep pace with demand.

Memory and Logic Customers Are Still Sold Out

ASML’s April investor call gave the Reuters interview more weight because the company had already described the same imbalance in unusually concrete terms. Fouquet said many memory customers had told ASML they were sold out for the rest of the year and expected supply limitations to last beyond 2026 despite plans to add significant capacity.

He said logic customers were also adding capacity across several advanced nodes while ramping 2-nanometer production for next-generation high-performance computing and mobile applications. In other words, the shortage is not confined to one corner of the semiconductor market. It is running through both memory and logic, which makes relief harder to achieve because both sides of the industry need tools, engineering talent, and factory time at once.

That helps explain why ASML said demand for chips is outpacing supply across markets ranging from AI to mobile devices and personal computers. When even more mature demand pools remain active while AI pushes the frontier higher, the cushion that normally absorbs short-term volatility becomes much thinner.

ASML Is Expanding Capacity as Chip Supply Pressures Grow

ASML is not presenting itself as a passive observer of the bottleneck. The company said in its quarterly update that order intake remains strong and that it is working with customers through a mix of new-system deliveries and productivity upgrades for the installed base.

That approach matters because the easiest way to loosen chip supply in the short term is not always building entirely new fabs. Sometimes the faster gain comes from pushing more wafers through existing tools, raising move rates, and bringing new generations of equipment into production quickly enough to support node transitions without slowing output elsewhere.

High NA EUV Is Moving From Promise to Production

One of the clearest signals from Reuters’ interview was Fouquet’s expectation that the first logic chips made with ASML’s High NA EUV systems would be produced within months. He also said product data for both logic and memory chips made using High NA tools should arrive this year, giving investors a nearer-term milestone for a technology that has long been treated as strategically important but commercially early.

That timing matters because High NA EUV is central to how the industry expects to keep shrinking features and improving performance in the next wave of advanced chips. If the first commercial outputs arrive on schedule, customers gain confidence that the lithography roadmap needed for future AI processors is moving from laboratory promise toward real manufacturing volume.

ASML also said Intel is set to be an early adopter, which reinforces the idea that High NA is not just a future option but part of active capacity planning today. In a supply-constrained market, the speed with which customers can qualify and scale new lithography tools has direct consequences for the revenue mix and margins of the broader chip ecosystem.

Productivity and Packaging Are Becoming New Chip Supply Levers

Fouquet said ASML is boosting output and improving the productivity of its tools as it tries to keep up with demand. On the April call, the company said it was executing on an output plan of at least 60 Low-NA EUV systems in 2026 and intended to lift Low-NA EUV capacity to at least 80 systems next year while scaling DUV and application products in parallel.

Those targets show that chip supply is no longer just a fab-construction story. It is also a tool-throughput story. If customers can get more wafers through the equipment they already own, and if ASML can lift deliveries without destabilizing the supply chain upstream, some of the pressure can be relieved before entirely new plants reach full qualification.

Reuters also reported that ASML is developing another advanced-packaging tool to help customers manufacture physically larger AI chips. That is important because chip performance is increasingly being improved not only by shrinking transistors but also by packaging multiple components together more efficiently. In practice, that makes packaging capacity part of the chip supply equation too.

Policy and Geography Will Shape the Next Chip Supply Phase

The Reuters interview did not stop at market demand. Fouquet also argued that policy choices in Europe and export-control disputes around China could influence how efficiently the next capacity cycle unfolds.

That is a useful reminder that semiconductor shortages are not driven only by technology and capex. Regulation, trade restrictions, and the pace of industrial adoption also affect where tools can go, how quickly new fabs scale, and which regions are best placed to capture the next wave of AI manufacturing value.

Europe Risks Slowing Its Own AI Opportunity

Fouquet said Europe risks being left behind in the industrial-adoption phase of AI because companies are often discouraged by the region’s regulatory complexity. He specifically argued that the European Union should scrap or revamp its AI Act, reflecting a view shared by some industrial leaders that heavy compliance requirements could slow deployment even when the technology base is available.

That criticism matters because Europe has been trying to present itself as both a trusted regulator and a serious industrial participant in AI infrastructure. ASML’s position gives that debate extra weight. The company is one of Europe’s most strategically important technology manufacturers, and its success depends on customers continuing to invest aggressively rather than waiting for regulatory clarity.

If Europe becomes a slower market for applied AI, some of the investment that might have fed regional semiconductor demand could shift elsewhere. That would not weaken ASML’s global role, but it could shape where the next clusters of compute capacity, packaging, and advanced manufacturing are built.

China Controls Still Hang Over the Chip Supply Outlook

Fouquet also called for more consistent rules around exports of chipmaking equipment to China after U.S. lawmakers proposed tighter allied restrictions that could extend to ASML’s lower-tech DUV tools. He said the DUV machines still sold into China are based on technology first introduced in 2015, underlining his argument that additional tightening may have limited strategic benefit while creating longer-term competitive effects.

His warning was that tougher restrictions could accelerate China’s efforts to develop domestic alternatives. That does not mean China can quickly replicate ASML’s leading position, but it does highlight a familiar pattern in industrial policy: the more durable the market barrier, the stronger the incentive for local substitution.

ASML said in its April outlook that its 2026 guidance range was wide enough to account for different outcomes from export-control discussions. That is a careful way of saying policy uncertainty has become part of normal planning. For the wider market, it means chip supply will be shaped not just by how many tools ASML can ship, but also by where governments decide those tools can be used.

Chip supply pressure is now one of the clearest signals of how deeply AI spending is changing the semiconductor industry, and ASML’s latest warning suggests the constraint is likely to last longer than many hoped. Keep reading related coverage at Berrit Media for more on AI infrastructure, semiconductors, and the policy shifts shaping global technology markets.