How ASML Built an Open-Innovation Supply-Chain Moat

1.The Product Required More Disciplines Than a Young Company Could Own

A lithography system brings together optics, light sources, precision motion, materials, sensing, software and process knowledge. Performance depends on their interaction.

For early ASML, attempting to build every specialist field internally would have expanded the task far beyond developing a competitive scanner.

Partnership was therefore a way to reach capabilities the system needed, not evidence that ASML contributed nothing itself.

The company's distinctive job was integration: decide how the pieces should work together and make the combined result useful to a semiconductor manufacturer. That responsibility remained with ASML even when another company supplied a critical subsystem. [S21][S22]

2.The ZEISS Relationship Established a Durable Pattern

ASML's formal relationship with ZEISS dates to 1986. Steef Wittekoek helped pioneer the partnership during the early platform period. [S21][S25]

Specialist optics could improve the scanner, while ASML supplied system requirements, integration and access to customer manufacturing problems.

That division created mutual opportunity. It also created dependence: each side needed the other to develop and deliver compatible capabilities.

The relationship was not a generic purchase order for an interchangeable lens. Its value grew as the partners accumulated knowledge about successive generations and the requirements at their interfaces.

3.Integration Became a Capability of Its Own

The best component in isolation is not always the best component in the complete system. A change can improve one metric while worsening weight, heat, reliability, lead time or cost elsewhere.

ASML had to choose among those trade-offs according to the customer's manufacturing objective.

This work is difficult to describe as one patent or machine part. It consists partly of architecture, specifications, testing and the experience needed to recognize where a subsystem improvement creates a system-level problem.

That is why “ASML buys many parts” does not explain away its engineering role. The customer needs a qualified scanner, not a collection of uncoordinated technical achievements.

4.A Roadmap Gave Specialist Investment a Commercial Direction

A supplier may need years of development and production preparation before a new scanner reaches customers. It cannot wait for broad volume orders before beginning every necessary investment.

ASML's roadmap can help organize those commitments by connecting future technical requirements with expected customer demand.

The forecast is not a guarantee. Semiconductor cycles and node timing can change, leaving a supplier with capacity or development expense ahead of revenue.

The commercial value of the relationship therefore depends on credibility over time. Partners need reason to believe that difficult work will lead to a usable system and a realistic manufacturing opportunity.

5.Customer Collaboration Reduced Another Kind of Uncertainty

Chipmakers help define what the next system must accomplish. Their process requirements, qualification work and production feedback prevent the roadmap from becoming only an equipment-maker's internal exercise.

The 2012 customer co-investment program illustrates a stronger version of that alignment: customers participated in arrangements that included R&D commitments before future technologies were fully commercial. [S64]

Participation did not remove engineering risk or make customers identical in their needs. It showed that equipment development and customer process roadmaps could require shared commitments.

The network therefore ran in both directions. Upstream specialists supplied capabilities, while downstream manufacturers supplied requirements, evidence and demand.

6.Cymer Showed That Partnership Was Not the Only Boundary Choice

ASML and Cymer collaborated before ASML announced the acquisition in 2012. The stated purpose of the transaction was closer coordination around EUV source development. [S56]

That decision did not contradict the partnership model. It showed that some interfaces had become so tightly linked to system performance that common ownership could improve development control.

The acquisition also preserved responsibilities to Cymer's existing DUV customers. Integrating a capability did not erase the acquired business's wider operating relationships.

ASML's approach was therefore selective: partner where specialist independence worked, and consider ownership where the roadmap required a different form of coordination.

7.A Minority Investment Was Different From a Full Acquisition

In 2016, ASML and ZEISS announced a 24.9% investment in Carl Zeiss SMT for €1 billion, alongside support for next-generation optics development and capacity. [S35]

The transaction did not make the whole ZEISS group an ASML subsidiary. Carl Zeiss SMT retained its place within the ZEISS organization.

That structure matched a particular relationship. It strengthened commitment to a shared technical roadmap without collapsing the partner into ASML.

The distinction matters because ownership percentages describe different levels of control. A strategic minority investment, a long-term supply agreement and a wholly owned subsidiary are not interchangeable ways of saying “ASML owns its supply chain.”

8.The EUV Source Still Relied on Contributions Across Companies

EUV source generation uses specialist laser technology as part of a broader physical system. TRUMPF's official description explains its role in supplying the laser technology that interacts with tin droplets to create the plasma. [S65]

Cymer's place inside ASML did not remove that external contribution. Nor did either contribution replace the need for optics, contamination control and scanner integration.

This is a useful example of nested responsibility. One internal capability can itself depend on an external specialist.

The competitive barrier is therefore distributed. Reproducing the machine requires more than copying the most visible subsystem or acquiring one supplier.

9.Berliner Glas Addressed Qualified Physical Capacity

ASML's 2020 acquisition of Berliner Glas strengthened a precision-module supplier relationship. Later disclosures connected the retained capability with the ramp-up and rollout of future systems. [S49]

The company then sold non-semiconductor activities. That sequence shows a focus on particular capabilities rather than indiscriminate expansion into every market served by an acquired group.

The transaction addressed a physical industrial problem: future systems needed modules that could be manufactured at the required quality and scale.

It also illustrates why ownership is not the same as immediate capacity. Integrating a supplier does not eliminate the time needed for processes, equipment, qualification and skilled people.

10.Manufacturing Qualification Made Replacement Difficult

A component that appears equivalent on paper may not behave identically within the scanner. Changing it can require testing, calibration or design adjustments elsewhere.

This makes qualified production knowledge valuable. The supplier knows how to make the part; ASML knows how it behaves in the system; customer experience confirms whether the combination works in a fab.

A competing network must reproduce those relationships, not merely obtain a component with a similar catalog specification.

That is a practical barrier to entry. It also means ASML may find it difficult to replace a disrupted supplier quickly. The same accumulated qualification that protects performance can reduce short-term flexibility.

11.More Demand Did Not Make Supply Immediately Expandable

The Q3 2026 release described customer forecasts exceeding ASML's then-planned supply for 2027 and 2028. Management was investigating further capacity additions. [S4]

That situation shows the difference between commercial demand and qualified industrial output.

Adding capacity involves many linked activities: supplier tooling, skilled work, module testing, integration space and customer installation. Expanding one stage can simply move the bottleneck to another.

The network's strength is the ability to coordinate these commitments. Its limit is that coordination cannot make every specialist process instantly scalable.

Qualified capacity is also different from the number of parts that leave a supplier's factory. A module may need testing, calibration and integration before it contributes to a finished scanner.

If a new production line has low yield or unstable quality, its nominal capacity can overstate the equipment output ASML can actually deliver. The next stage may then spend time diagnosing modules rather than assembling systems.

This makes manufacturing knowledge part of the relationship. Suppliers and ASML need to understand not only the design but the repeatable process for producing it.

It also explains why distributing a difficult task among more suppliers is not always an immediate solution. Additional sources may improve resilience in the long term, but each must establish the relevant capability and pass qualification. A diversification plan can require substantial work before it reduces a current bottleneck.

12.Field Knowledge Extended the Network Into Customer Operations

After installation, support engineers encounter the equipment under real production conditions. They learn which failures recur, which improvements are practical and how configurations behave over time. [S51]

That information can feed back into product and supplier development. A component change may be evaluated not only against factory tests but against experience from a deployed fleet.

This adds another layer of knowledge that is difficult to recreate quickly. It is operational, distributed and accumulated over many customer interactions.

The company must handle it carefully. Confidential process information and differences among applications limit what can be generalized or shared across the network.

13.Open Innovation Did Not Mean Open Access to Every Secret

The term describes collaboration among organizations with complementary capabilities. It does not imply that all designs, recipes or intellectual property are public.

The partners need appropriate information to solve shared technical problems while preserving the boundaries required by contracts, customer trust and their own businesses.

This makes the operating model more demanding than either a fully internal hierarchy or a simple spot market.

Too little coordination can delay the system. Poorly managed disclosure can damage the relationships that make coordination possible. The network's value depends on both connection and boundaries.

14.The Moat and the Risk Came From the Same Structure

Specialization makes the combined capability hard to replicate. It also creates concentrated dependencies.

A supplier interruption can hold back the whole scanner. A customer delay can change the economic basis for upstream investment. A regulatory change can affect where knowledge or equipment can move.

ASML's strong position therefore does not mean independence from others. It means the company occupies a valuable coordinating role among parties that also depend on one another.

The enduring test is whether that organization can deliver the next qualified system as conditions change—not whether it can claim to manufacture every part.

There is a corresponding limit to financial leverage. A stronger order forecast can justify investment, but it cannot instantly create experienced technicians or a proven precision-manufacturing process.

Capital is an input to expansion. The usable result arrives only when the people, equipment, quality and interfaces work together. That delay is part of the network's economic structure, not merely a temporary scheduling inconvenience.