How Philips Accidentally Created Europe's Most Strategic Company

In 1984, Philips spun off a small lithography unit. Today ASML controls the global semiconductor supply chain. The most productive corporate breakup in European history.

By VastBlue Editorial · 2026-03-26 · 24 min read

Series: How We Got Here · Episode 5

The Conglomerate That Couldn't Hold

By the early 1980s, Royal Philips Electronics was one of the largest companies in the world and one of the most confused. Headquartered in Eindhoven, a city in the southern Netherlands whose entire identity had been shaped by the company for nearly a century, Philips manufactured lightbulbs, televisions, radios, cassette players, semiconductors, medical imaging equipment, shavers, defence electronics, and — tucked away in a division that almost nobody outside the company had heard of — machines that printed patterns onto silicon wafers. The company employed over 340,000 people across more than sixty countries. Its revenues exceeded $30 billion. Its organisational chart resembled the wiring diagram of one of its own products: dense, recursive, and comprehensible only to specialists.

This was not an accident of growth. It was a deliberate strategy — or rather, the accumulated residue of decades of deliberate strategies, each sensible in its moment, collectively producing an entity that did too many things to do any of them with the focus required by increasingly competitive global markets. Philips had been a conglomerate since the 1930s, when it began diversifying beyond lightbulbs into radio technology. By the 1960s and 1970s, when conglomerates were fashionable and management theory held that a well-run company could manage any business, Philips expanded into virtually every corner of the electronics industry. The logic was vertical integration: control the components, control the assembly, control the distribution, control the consumer relationship. From silicon to storefront.

The problem was that by the 1980s, each of these markets was becoming a war in its own right. Japanese manufacturers — Sony, Matsushita, Toshiba, Sharp — were eating into Philips' consumer electronics business with better products at lower prices. American companies — Intel, Texas Instruments, Motorola — were pulling ahead in semiconductors. Siemens and General Electric were competitive in medical equipment. In each market, Philips was present but rarely dominant, investing enough to stay in the game but never enough to win it. The conglomerate structure, which had once provided the advantages of scale and integration, was now a tax on focus. Engineering talent was spread across dozens of product lines. Capital was allocated through internal political negotiations rather than market signals. The cross-subsidisation that had once allowed promising divisions to grow now meant that profitable units funded the losses of uncompetitive ones.

The company's semiconductor lithography operation exemplified this dynamic. Philips had been working on lithography — the process of using light to transfer circuit patterns onto silicon wafers — since the early 1960s, when its Natlab research laboratory in Eindhoven began developing optical systems for the emerging semiconductor industry. The work was technically excellent. Philips' optical engineers were among the best in the world, inheritors of a tradition that stretched back to the company's origins in precision lamp manufacturing. They understood light — how to generate it, focus it, control it — at a level that few organisations could match. But inside Philips, the lithography unit was marginal. It was a tiny operation generating modest revenue in a company dominated by consumer electronics and lighting. It competed for capital and attention against divisions that were larger, more profitable, and more politically connected within the organisation. The engineers working on lithography knew they had something important. The corporate hierarchy was not listening.

The Birth of ASML

In 1984, Philips did something that looked, from the outside, like a routine piece of corporate housekeeping. It spun off its lithography division into a joint venture with ASM International, a Dutch semiconductor equipment company. The new entity was called ASM Lithography — ASML. It was headquartered not in Eindhoven proper but in Veldhoven, a suburb so quiet that its main distinction was proximity to Philips' campus. The joint venture had around 100 employees. Its first-year revenue was approximately $13 million. In the context of Philips' $30 billion empire, this was a rounding error.

The circumstances of the spinoff were not glamorous. Philips was under financial pressure. Wisse Dekker, who had become CEO in 1982, was attempting to rationalise a sprawling company that had become unprofitable. Divisions that were not core to the consumer electronics business — which Dekker and his successor Cor van der Klugt saw as Philips' future — were candidates for divestiture, joint venture, or closure. The lithography operation fell into the joint-venture category: too small to matter, too technically interesting to kill, and potentially viable if it could find a partner to share the development costs.

ASM International, run by Arthur del Prado, a Dutch entrepreneur who had built the company into a significant player in semiconductor manufacturing equipment, provided the commercial half of the equation. Philips provided the optical technology, the engineering talent, and the legitimacy of a major corporate name. The joint venture was structured as a 50-50 partnership, with both parents contributing technology and the expectation that ASML would either prove itself commercially or be wound down within a few years.

ASML's early years were precarious. The semiconductor lithography market in the mid-1980s was dominated by two companies: GCA Corporation of the United States and Nikon of Japan. Canon was emerging as a third competitor. ASML was a distant fourth, with inferior market share, limited customer relationships, and a product — the PAS 2500 stepper — that was technically promising but commercially unproven. The company lost money in its first several years. Its survival was uncertain enough that Philips and ASM International periodically debated whether to continue funding it.

What saved ASML was a combination of technical insight and institutional design that, in retrospect, looks almost impossibly fortunate. The technical insight was that semiconductor lithography would become the bottleneck of Moore's Law — that the ability to print ever-smaller patterns on silicon wafers would determine the pace of the entire semiconductor industry's advance. As chip features shrank from micrometres to fractions of micrometres, the optical systems required to print them became exponentially more complex. This was not a market where good-enough products could survive. It was a market where only the most advanced technology would be purchased, because chipmakers buying a lithography system were making a bet on their ability to manufacture the next generation of chips. Second-best lithography meant second-best chips, which meant no customers.

The Architecture of Dominance

ASML's rise from marginal startup to monopoly controller of semiconductor lithography is one of the most remarkable stories in the history of industrial technology. It happened not through a single breakthrough but through a sustained, decades-long accumulation of technical capability, strategic partnerships, and institutional architecture that, piece by piece, made ASML indispensable.

The first critical decision was architectural. In the late 1980s, ASML adopted a modular design philosophy for its lithography systems. Rather than building every component in-house — the approach favoured by Nikon and Canon, both vertically integrated Japanese manufacturers — ASML designed its systems as platforms that integrated best-in-class components from specialised suppliers. The optical column came from Carl Zeiss in Germany. The laser light source came from Cymer in the United States (which ASML would eventually acquire in 2013). The wafer stages, mirrors, and control systems were sourced from a network of specialist firms, each pushed to the frontier of their particular discipline by ASML's relentless performance requirements.

This modular approach gave ASML three advantages that compounded over time. First, it allowed the company to focus its internal engineering resources on systems integration — the extraordinarily difficult task of making all these precision components work together to print features measured in nanometres. Second, it created a supply chain in which each supplier was investing its own R&D capital in pushing the boundaries of its specialty, effectively multiplying ASML's total innovation capacity far beyond what any single company could achieve. Third, it made ASML's suppliers dependent on ASML as their primary customer, creating a gravitational pull that drew the best component technology toward ASML's platform rather than toward competitors.

The second critical decision was the bet on extreme ultraviolet (EUV) lithography. By the mid-1990s, the semiconductor industry was approaching the physical limits of deep ultraviolet (DUV) lithography — the technology that had driven chip miniaturisation for two decades. Features were shrinking toward the wavelength of the light used to print them, a fundamental physical barrier. Multiple alternatives were proposed: X-ray lithography, electron beam lithography, ion beam lithography, and EUV lithography, which would use light with a wavelength of 13.5 nanometres — roughly fourteen times shorter than the DUV wavelength of 193 nanometres. Each approach had advocates and sceptics. None was proven.

ASML bet on EUV. The bet was enormous — both financially and technically. EUV lithography required reinventing virtually every aspect of the lithography system. At 13.5 nanometres, light is absorbed by air and by glass, which meant the entire optical path had to operate in a near-perfect vacuum and use reflective mirrors rather than refractive lenses. The mirrors — manufactured by Carl Zeiss SMT with surfaces polished to an accuracy measured in individual atoms — were the most precise optical components ever manufactured. The light source — a system that fires a high-powered laser at tiny droplets of molten tin, producing a plasma that emits EUV radiation at 50,000 droplets per second — had no precedent in industrial manufacturing. Each EUV system contains more than 100,000 parts, weighs approximately 180 tonnes, and requires multiple Boeing 747 cargo planes to ship.

An EUV lithography system contains more than 100,000 parts, weighs approximately 180 tonnes, and requires multiple Boeing 747 cargo planes to deliver. Each mirror is polished to an accuracy measured in individual atoms. This is engineering at the absolute boundary of what physics permits.

Based on ASML technical disclosures

The EUV programme took more than twenty years from inception to commercial deployment. ASML shipped its first commercial EUV system in 2010, but it took until 2019 for the technology to reach the volume production readiness that chipmakers required. The total investment — by ASML, its suppliers, its research partners, and the chipmaker customers who co-funded development — exceeded $10 billion over the life of the programme. There were years when the project appeared likely to fail. The light source, in particular, was a seemingly intractable problem: early prototypes produced so little EUV radiation that exposing a single wafer took hours rather than the seconds required for commercial viability. The solution required breakthroughs in laser engineering, tin droplet generation, plasma physics, and debris mitigation that pushed multiple scientific disciplines simultaneously.

When EUV finally worked, it created a monopoly. No other company — not Nikon, not Canon, not any Chinese or American firm — had the combination of optical technology, light source engineering, systems integration capability, and supplier relationships required to build an EUV lithography system. ASML's market share in EUV is not 90 per cent or 95 per cent. It is 100 per cent. Every advanced semiconductor manufactured anywhere in the world — every chip in every iPhone, every GPU training AI models, every processor in every data centre — is printed using an ASML machine. There is no alternative supplier. There is no substitute technology. This is not a market. It is a chokepoint.

The Philips Diaspora

ASML is the most consequential offspring of Philips' long corporate fragmentation, but it is not the only one. Between the 1980s and the 2020s, Philips systematically divested, spun off, or sold virtually every business it had built over the previous century. The result is a constellation of independent companies — several of them global leaders in their respective markets — that trace their DNA directly to the Eindhoven conglomerate.

NXP Semiconductors, spun off from Philips' semiconductor division in 2006, is the most significant after ASML. Philips had been manufacturing semiconductors since the 1950s, and by the early 2000s its semiconductor arm was one of the largest in Europe, with particular strength in automotive chips, near-field communication (NFC) technology, and secure identification. The spinoff was executed as a leveraged buyout led by private equity firms KKR, Bain Capital, and Silver Lake, with Philips retaining a minority stake. NXP went public in 2010 and subsequently acquired Freescale Semiconductor in 2015, creating a company with dominant positions in automotive semiconductors and secure connectivity. Today NXP is worth approximately $50 billion and supplies chips to virtually every major automobile manufacturer in the world. Its NFC technology is embedded in billions of contactless payment cards and mobile devices.

Signify, spun off in 2016, inherited Philips' original business: lighting. The company that Gerard Philips founded in 1891 to manufacture incandescent lightbulbs had evolved through a century of lighting technology — fluorescent tubes, halogen lamps, compact fluorescents, and finally LED systems. Signify retained the Philips brand under licence and is now the world's largest lighting company, with particular strength in connected lighting systems and horticultural lighting for controlled-environment agriculture. The irony is structural: the business that created Philips became an independent company that pays Philips for the right to use its name.

Versuni, created in 2023, took Philips' domestic appliances business — coffee machines, air fryers, vacuum cleaners, garment steamers — and became an independent company backed by the investment firm Hillhouse. TP Vision, a joint venture with the Taiwanese manufacturer TPV Technology, took over Philips' television business. Philips' audio and video operations were sold to various acquirers over the years. The defence electronics business was sold to Thomson-CSF (now Thales). The recorded music business — Philips had co-invented the compact disc with Sony and owned PolyGram, one of the world's largest record labels — was sold to Seagram in 1998 and eventually became part of Universal Music Group.

• ASML (1984) — Semiconductor lithography. Market cap: ~€260 billion. 100% of EUV market.
• NXP Semiconductors (2006) — Automotive and secure chips. Market cap: ~$50 billion. Global leader in automotive semiconductors.
• Signify (2016) — Lighting. Revenue: ~€6.9 billion. World's largest lighting company.
• Versuni (2023) — Domestic appliances. Major global home appliances player.
• TP Vision — Consumer televisions. Operates Philips TV brand globally.
• PolyGram/Universal Music Group — Recorded music. Sold 1998. Now world's largest music company.

What remains of Philips today is a healthcare technology company — MRI scanners, CT scanners, ultrasound systems, patient monitoring equipment, health informatics. This is itself a legacy of the conglomerate era: Philips entered medical imaging in the 1890s with X-ray equipment, barely a year after Wilhelm Röntgen's discovery, and built a formidable position in diagnostic imaging over the following century. The current Philips, with revenues of roughly €18 billion and a market capitalisation that fluctuates around €25 billion, is a focused, mid-sized medical technology company. It is worth less than a tenth of ASML.

The Geopolitics of a Chokepoint

ASML's monopoly on EUV lithography did not become geopolitically significant until the United States and China began treating semiconductor manufacturing as a theatre of strategic competition. That happened decisively in October 2022, when the US Bureau of Industry and Security issued export controls restricting China's access to advanced semiconductor manufacturing equipment. The controls were expanded in October 2023 and again in 2024. Their central target, though never explicitly named in the regulatory text, was ASML.

The logic was straightforward. Advanced semiconductors — the chips required for artificial intelligence training, advanced weapons systems, supercomputing, and next-generation telecommunications — can only be manufactured using EUV lithography. EUV lithography can only be obtained from ASML. If ASML cannot sell to China, China cannot manufacture advanced chips. The chokepoint is not a tariff, a quota, or a negotiating position. It is a physical reality: no EUV machine, no advanced chips. There is no workaround. China has been investing heavily in domestic semiconductor equipment development, but the consensus among industry experts is that replicating ASML's EUV capability would require a decade or more, assuming the underlying technical problems — which took ASML twenty years and tens of billions of dollars to solve with the full cooperation of the world's best optical and laser engineers — could be solved at all.

The Netherlands — a country of 17 million people, famous for tulips, canals, and pragmatic consensus politics — found itself holding the most consequential export-control lever in the semiconductor supply chain. The Dutch government, under intense pressure from Washington, agreed in January 2023 to restrict exports of ASML's most advanced DUV systems as well as all EUV systems to China. The decision was not made lightly. ASML is the Netherlands' most valuable company and one of its largest employers. China was ASML's fastest-growing market, accounting for roughly 15-20 per cent of revenue. Restricting sales to China meant sacrificing billions in revenue and risking retaliation against Dutch commercial interests in the Chinese market.

A small country in northwestern Europe, through an improbable sequence of corporate decisions, technological bets, and institutional accidents, ended up holding one of the most strategically significant positions in the global technology order. Nobody planned it. The consequences are enormous.

Editorial observation

The geopolitical implications extend beyond US-China competition. ASML's position means that Europe — often dismissed in technology discourse as a consumer of American and Asian innovation rather than a producer — controls a chokepoint without which neither American chip designers nor Asian chip manufacturers can function. Intel, TSMC, Samsung — the three companies building the world's most advanced semiconductor fabrication plants — are all dependent on ASML equipment. This gives Europe, and specifically the Netherlands, a form of strategic leverage that no amount of industrial policy could have deliberately created. It was not planned. It was the downstream consequence of a corporate spinoff executed four decades ago by managers who were trying to clean up a bloated conglomerate.

ASML itself navigates this terrain with the careful pragmatism characteristic of Dutch business culture. The company complies with Dutch and European export regulations, advocates publicly for the broadest possible market access — "We sell to anyone who can legally buy," as CEO Peter Wennink repeatedly stated before his retirement in 2024 — and invests heavily in next-generation technology to maintain the technical lead that underpins its position. The company's current development programme, High-NA EUV lithography, pushes the numerical aperture of the EUV optical system to 0.55 (from 0.33 in current systems), enabling chip features below two nanometres. The first High-NA system was shipped to Intel in late 2023. Each system costs approximately $380 million. There is, again, no alternative supplier.

What the Philips Story Teaches

The Philips diaspora is the most instructive corporate breakup in European industrial history. Not because it was well-executed — it was messy, reluctant, and driven more by financial pressure than strategic vision — but because of what it reveals about how strategic industrial capability is actually created.

The first lesson is that conglomerates suppress the value of their best assets. ASML inside Philips was a marginal division competing for attention against businesses that were larger, more established, and more politically connected within the corporate hierarchy. ASML outside Philips was free to pursue a singular technical vision with the intensity and focus required to achieve it. The same pattern repeated with NXP: Philips' semiconductor division, starved of investment within the conglomerate, became a globally competitive force once it was liberated from the parent company's internal capital allocation politics. The lesson is not that conglomerates are always wrong. It is that when a business requires deep, sustained, technically ambitious investment over decades — as semiconductor lithography and semiconductor manufacturing both do — the conglomerate structure is precisely wrong because it optimises for diversification, risk management, and internal political balance, not for the maniacal focus that frontier technology demands.

The second lesson is that strategic industries are not created by strategic planning. No Dutch policymaker in 1984 sat down and decided that the Netherlands should control the global semiconductor lithography supply chain. No European Commission white paper identified photolithography as a strategic capability to be nurtured. ASML's dominance emerged from a sequence of technical decisions, commercial bets, and institutional accidents that, taken individually, were responses to immediate problems rather than elements of a grand design. The decision to adopt a modular architecture was a response to ASML's lack of scale: it could not afford to develop every component in-house, so it built a platform. The bet on EUV was a response to the physical limits of DUV lithography: the laws of physics, not a strategy document, dictated the direction. The acquisition of Cymer was a response to supply-chain vulnerability: ASML needed to control the light source, so it bought the company that made it.

The third lesson is that once a chokepoint exists, it becomes geopolitical whether its creators intended it or not. ASML's engineers were trying to build the world's best lithography machines. They were not trying to create a lever of geopolitical power. But the nature of their achievement — a capability so advanced that no competitor can replicate it within any policy-relevant timeframe — made geopolitics unavoidable. The company that started as a corporate spinoff employing 100 people in a Dutch suburb now sits at the intersection of the US-China technology competition, European industrial sovereignty debates, and the global race to secure semiconductor supply chains. Its export decisions are discussed in the White House. Its technology roadmap is studied by intelligence agencies. Its hiring plans affect the strategic calculations of multiple governments.

The fourth lesson concerns patience — specifically, the kind of patient, sustained, technically ambitious investment that ASML's EUV programme represents. Twenty years from concept to commercial deployment. More than $10 billion in cumulative investment. Repeated near-death experiences as technical obstacles appeared insurmountable. No quarterly earnings report, no venture capital fund, no public market equity analyst would have tolerated this timeline. EUV survived because ASML's customers — Intel, TSMC, Samsung — co-funded the development programme, because they understood that without EUV their own roadmaps would hit a wall. This collaborative, long-horizon investment model has no equivalent in American venture capital or European public funding. It is a sui generis institutional achievement, born of necessity, and it is the reason ASML has no competitors.

Europe in the mid-2020s is engaged in a frantic effort to build strategic capability in semiconductors, batteries, AI, and other frontier technologies. The European Chips Act allocates billions. National governments announce fab construction projects. Policy documents invoke "strategic autonomy" and "technological sovereignty." These efforts are well-intentioned and, in many cases, necessary. But the Philips story suggests that the most consequential strategic capabilities may emerge not from top-down planning but from the messy, contingent, often accidental processes of corporate evolution — and that the role of government is less to direct the creation of strategic industries than to recognise them when they appear and create the conditions in which they can grow.

ASML exists because Philips could not hold itself together. It dominates because its engineers made a series of technically correct bets over four decades. It matters geopolitically because the capability it created has no substitute. Every element of this story was contingent. None of it was inevitable. And the small city of Veldhoven — population 45,000, surrounded by farms and light industrial parks in the southern Netherlands — is now one of the most strategically significant locations on Earth. If you had told the Philips board in 1984 that their most lasting contribution to European industry would be a 100-person spinoff making machines that nobody outside the semiconductor industry had heard of, they would not have believed you. History has a sense of irony that corporate strategy rarely accounts for.

Sources

1. Rene Raaijmakers. "ASML's Architects." Techwatch Books, 2014. — https://www.techwatch.nl/asmls-architects/
2. Chris Miller. "Chip War: The Fight for the World's Most Critical Technology." Scribner, 2022. — https://www.simonandschuster.com/books/Chip-War/Chris-Miller/9781982172008
3. ASML Annual Report 2023. — https://www.asml.com/en/investors/annual-report
4. US Bureau of Industry and Security. "Implementation of Additional Export Controls: Certain Advanced Computing and Semiconductor Manufacturing Items." Federal Register, October 2022. — https://www.federalregister.gov/documents/2022/10/13/2022-21658/
5. Philips Annual Report 2005 (pre-NXP spinoff). — https://www.philips.com/a-w/about/investor-relations/annual-reports.html
6. National Academies of Sciences. "Securing the Future of the U.S. Semiconductor Industry." 2023. — https://nap.nationalacademies.org/
7. European Commission. "European Chips Act." Regulation (EU) 2023/1781. — https://digital-strategy.ec.europa.eu/en/policies/european-chips-act
8. Varas, Antonio, et al. "Strengthening the Global Semiconductor Supply Chain in an Uncertain Era." Boston Consulting Group and Semiconductor Industry Association, 2021. — https://www.bcg.com/publications/2021/strengthening-the-global-semiconductor-supply-chain