The lithography machine giant has collapsed

Recently, Japanese optical giant Nikon issued its most severe loss warning in history—expecting a massive loss of 85 billion yen in the fiscal year 2025, marking the worst record for the company since its founding in 1917. Its core lithography business has suffered a complete collapse, putting the once-dominant lithography leader in an unprecedented survival crisis. According to multiple reports, in the past six months, Nikon shipped only 9 lithography machines, all of which were older models with lower technical content, clearly lagging in technological generations. This means that this company, which was once deeply tied to Intel and AMD and set industry standards, has completely lost its competitiveness in the advanced process field. Nikon not only failed to capitalize on the boom in AI computing power but also fell into an unprecedented financial quagmire due to a significant shrinkage in orders and inventory backlog. In stark contrast, in 2025, Dutch ASML sold 327 units, with 48 units of high-end EUV lithography machines, dominating the global high-end market. A Turning Point of an Era Once known as one of the "three giants of lithography machines" alongside ASML and Canon, Nikon held about 40% of the global lithography machine market share in 2001, with nearly one in every two lithography machines produced by Nikon. This company, once sought after by chip giants, has seen its market share drop to single digits, with its market competitiveness nearly zero. The fall of Nikon from peak to trough did not happen overnight. Its fate reflects the changes in the global lithography machine market over the past thirty years and poses a brutal question to the industry: When the industry leader is far ahead, how should the former king position itself? When the technological route is locked by competitors, do latecomers still have a chance to turn the tables? The Path of Nikon's Collapse from Peak to Trough The "Golden Age" of Nikon Lithography Machines To understand Nikon's fall, one must first return to its glorious years. Nikon's lithography machine business began in the 1970s, leveraging its core technological advantages in camera lenses to quickly enter the semiconductor lithography equipment market. At that time, the global semiconductor industry was in the early stages of rapid rise, with chip processes gradually transitioning from micrometers to nanometers. Lithography equipment, as the most core and complex device in chip manufacturing, became the focus of competition among major companies. With precise market judgment and leading optical technology, Nikon quickly established a foothold in the lithography machine market and experienced an explosion in the 1980s. At that time, Nikon launched a 193nm wavelength dry lithography machine, which became the preferred equipment for global chip manufacturers due to its ultra-high resolution and stability, dominating the market of the 193nm dry lithography era. According to industry data, in the mid-1990s, Nikon's global market share of lithography machines once exceeded 50%, evenly matched with Canon, with the two companies collectively occupying over 90% of the global lithography machine market, forming a monopoly pattern of "the two Japanese giants." During this period, Nikon's core competitiveness lay in its deep ties with the world's top chip companies At that time, American chip giants like Intel and AMD were fully promoting the upgrade of CPU processes, and Nikon's lithography machines, with their stable performance and leading technology, became core suppliers for these companies. The lithography equipment customized by Nikon for Intel perfectly matched its CPU production needs, helping Intel gain an advantage in the competition with AMD. It is reported that global chip giants, from Intel, AMD, IBM to Texas Instruments, established dedicated teams to stay at Nikon's Silicon Valley branch just to secure priority supply rights for a Nikon lithography machine. There are even rumors that some semiconductor executives personally visited Nikon's factory to wait, prepaying the full amount just to secure a debugging slot. This deep binding also allowed Nikon to obtain stable orders and substantial profits, further consolidating its industry position. In addition to binding with American giants, Nikon also has a strong customer base in Japan. Japanese semiconductor companies such as Sony, Toshiba, and Hitachi are all core customers of Nikon, and this advantage of "local synergy" has given Nikon a significant boost in the global market. At its peak, Nikon's lithography machines were not only a benchmark of technology but also the standard-setters of the industry, with its lithography technology specifications adopted by most chip manufacturers worldwide. Under Nikon's dominance, the American lithography machine pioneer GCA was forced to declare bankruptcy; at that time, ASML was still just a struggling small manufacturer in the European market with a market share of less than 10%, and could not be compared to Nikon at all. At that time, Nikon was indeed in a glorious position, even surpassing today's ASML in terms of brilliance. The lithography business became the core profit pillar of the group, driving the development of other businesses such as cameras and telescopes. Nikon also once became a pride of Japanese manufacturing, regarded as a model of a technology-driven nation. No one would have thought that such a king standing at the top of the industry would find itself in such a difficult situation just two or three decades later. Three major mistakes, step by step missing the wave of the era The turning point occurred in 2002. That year, Lin Benjian, then a senior director at TSMC, knocked on Nikon's door. In response to the bottleneck faced by the 193nm dry lithography machine and the slow progress of the next-generation 157nm light source, Lin proposed a disruptive idea: injecting a layer of water between the lens and the wafer. By utilizing the refractive index of water, the equivalent wavelength of the 193nm light source could be shortened to 134nm, thereby bypassing many challenges of the 157nm route. This was the immersion lithography technology route that later changed the history of semiconductors. This was originally a shortcut with lower costs and better effects, but it was met with opposition from almost all of Nikon's executives. From the chairman to the technical leaders, no one had the patience to listen to Lin's explanation. Nikon's representatives questioned on the spot: "If the water contaminates the lens, can TSMC afford the compensation? If bubbles cause batch scrapping, who will take responsibility?" The deeper reason lies in path dependence. At that time, Nikon had already invested over hundreds of millions of dollars in the 157nm dry lithography machine. Switching to the immersion route meant that all previous investments would be wasted According to Huashang Taolue reports: Nikon not only rejected Lin Benjian but even attempted to use its industry prestige to suppress this idea. Lin Benjian later recalled that Nikon's senior management had called TSMC's R&D Vice President Jiang Shangyi, saying: "Please take care of your Lin Benjian and do not let him promote this disruptive idea that undermines industry consensus, as it will distract everyone and waste resources." After hitting a wall with Nikon, Lin Benjian flew to the Netherlands. At that time, ASML was still struggling to survive and urgently needed a breakthrough opportunity. Martin van den Brink, the technical soul of ASML, went against the consensus and bet all of ASML's resources on this crazy idea. In 2004, ASML collaborated with TSMC to launch the world's first immersion lithography machine, ArFi, which swept the global market with higher precision and lower costs. By 2007, ASML's market share exceeded 60%, forming a crushing momentum for the first time; after 2010, ASML's market share surpassed 70%, completely widening the gap with Nikon and Canon. Nikon's proud top lenses instantly lost their luster in the face of the new technological route. Nikon and Canon were forced to abandon the 157nm route and follow immersion technology, but it was already too late. In the field of immersion ArF lithography, ASML had firmly grasped over 90% of the market share with its mature TWINSCAN dual-stage technology. This was a textbook-level technological misjudgment. Nikon was not lacking in technical capability, but was trapped by its own successful experience, exhibiting a natural rejection of new technologies outside its system. However, the failure of immersion was just the beginning; Nikon's real "Waterloo" was yet to come. In the face of the disastrous defeat in the immersion lithography battle, Nikon pinned its hopes on the next-generation technology: EUV (Extreme Ultraviolet Lithography). This technology, with a shorter wavelength (13.5nm) capable of etching smaller circuits on chips, was seen as the key battle for its return to the peak. At that time, Nikon's lithography technology head, Masahiro Matsumoto, set ambitious goals: fully self-developed and entirely produced in Japan. He attempted to replicate the era of precision manufacturing that conquered the world within the closed walls. At the same time, the Japanese government, having lost its chip dominance, also fully supported this, viewing it as a battle for national fortune. Led by the Ministry of Economy, Trade and Industry, Japan built a large "industry-government-academia" consortium, investing hundreds of billions of yen, uniting industry chain companies such as Nikon, Canon, Tokyo Electron, and Shin-Etsu Chemical to tackle the challenges together. This was a typical Japanese-style charge: concentrated resources, singular goals. But by this time, the world had changed. In 2012, when Nikon was fully committed to the EUV project, ASML received its first large-scale strategic investments from Intel, Samsung, and TSMC. The three major customers jointly funded ASML to accelerate the development of EUV while establishing their own EUV alliance. This alliance not only bound the world's top chip manufacturers but also gathered the strongest industry chain companies globally, such as Zeiss (lenses) from Germany and Cymer (light sources) from the United States This "vertical cooperation" model allows ASML to concentrate resources on system integration and core technological breakthroughs, rather than being all-encompassing. This is also one of the deep reasons for Nikon's failure. For a long time, Japanese companies have firmly believed in a fully self-researched production model, choosing to highly self-develop core components (lenses, light sources, precision machinery). This "vertical integration" can ensure extreme quality in an era of slow technological iteration, but when the era of EUV, which requires global high-level collaboration, arrives—with research and development costs easily exceeding tens of billions of dollars and involving 100,000 components—Nikon found that it could no longer afford this entry ticket. ASML's choice of "bundled interests and shared risks" led it down a completely different path. Worse still, the United States, which had suffered significant losses from Japanese companies in the chip sector, excluded Japanese manufacturers like Nikon and Canon from the EUV technology alliance on the grounds of national security, cutting off their access to top American technology. At this point, Nikon's "fully self-researched" approach turned into "reinventing the wheel behind closed doors." By 2018, Nikon's investment in the EUV project was estimated to exceed 100 billion yen, making it the largest single technological bet in the company's history. However, this investment resulted in only a non-commercial prototype. While ASML's EUV lithography machines had already been rapidly iterating on TSMC's production lines, Nikon's prototype was still gathering dust in the laboratory. When TSMC announced mass production of the 7nm process in 2018, ASML monopolized 90% of the global high-end lithography machine orders with EUV, forming a technological hegemony with no substitutes. Ultimately, Nikon had to announce the termination of the commercialization development of EUV lithography machines. In addition to a series of misjudgments in its technological route, Nikon also made fatal errors in its market strategy. It overly bet on a single giant, Intel. In 2024, Intel significantly reduced capital expenditures due to massive losses, directly leading to a sharp drop in Nikon's orders. At the same time, Nikon failed to timely expand its reach to core chip manufacturers like TSMC and Samsung, leaving a gap in orders that could not be filled. The external policy environment further exacerbated the situation. Over the past five years, China had been Nikon's largest "lifeline." With the expansion of mainland wafer fabs, Nikon's precision equipment sales to China once accounted for over 40%. However, when the U.S. implemented export controls on semiconductor equipment to China, Nikon chose to closely follow the U.S. lead, abandoning cooperation opportunities, which led to delays in equipment delivery and soaring costs. Chinese customers turned to domestic alternatives, further squeezing its survival space. Nikkei Asia pointed out that China has become the third country in the world with complete lithography machine manufacturing capabilities, and Nikon has long missed the opportunity to share in the market with its high-priced old models. In September 2025, Nikon closed its Yokohama factory, which had been in operation for 58 years, marking a further contraction of its lithography machine business. Meanwhile, 70-year-old Masahiro Ma, who is about to step down, has tried to restore past glory single-handedly, but ultimately found it beyond his capabilities. ASML: From "Defensive" to "Offensive" As Nikon gradually moved towards collapse, ASML transformed from an industry follower into the absolute leader in the global lithography machine market In the field of high-end lithography machines, ASML's monopoly position is unmatched. Especially in the EUV lithography market, ASML stands alone, controlling the "throat" of advanced process chip manufacturing at 7nm and below. Whether it's TSMC, Samsung, or Intel, they all rely on ASML's EUV lithography machines. According to statistics, ASML holds a 100% market share in the EUV lithography machine market and over 90% in the high-end DUV lithography machine market, forming a solid technological barrier and market moat. This constitutes its "cash cow" and monopoly foundation. But ASML has not stopped there. As Moore's Law approaches physical limits, the cost and difficulty of improving chip performance solely through transistor miniaturization are increasing. The industry is turning its attention to another direction: advanced packaging. As chip processes continue to approach physical limits, advanced packaging technology has become an important path to enhance chip performance. This is precisely the core technology relied upon by NVIDIA's AI chips like H100/B200—highlighting the importance of TSMC's CoWoS, InFO, and other packaging technologies. ASML has keenly realized: merely controlling "front-end manufacturing" may no longer be sufficient to dominate the future. If it can gain an advantage in the advanced packaging equipment field, it can extend its control from "front-end manufacturing" to "back-end packaging," achieving mastery over the entire chip manufacturing process, further expanding its market share and consolidating its industry dominance. Thus, ASML has shifted to "offensive" mode, beginning to layout and explore the advanced packaging equipment field. It can be said that while competitors are still mired in the mud, the winner has already begun to redefine the new battlefield. In October 2025, ASML took a substantial step by launching its first advanced packaging lithography machine, TWINSCAN XT:260, officially entering the advanced packaging market. This device uses a 365nm i-line light source to achieve 400nm resolution patterning, primarily applied in key processes such as RDL and TSV. Its overlay accuracy reaches ±1.2nm, a 52% improvement over the previous generation, with a production efficiency of 270 wafers per hour, a fourfold increase compared to the previous generation. The launch of TWINSCAN XT:260 marks ASML's official entry into the advanced packaging equipment market, and leveraging its technological advantages and brand influence accumulated in the lithography equipment field, it quickly gained market recognition. According to industry sources, core customers like TSMC and Samsung have already placed orders for ASML's advanced packaging lithography machines for their Chiplet technology research and mass production. But this may just be the beginning. Recently, industry insiders revealed that ASML has begun developing hybrid bonding machines and is collaborating with partners such as Prodrive and VDL-ETG, suppliers of magnetic levitation system components for EUV lithography machines. Hybrid bonding is the core technology for the next generation of 3D integration, enabling direct bonding of copper to copper, eliminating bumps and significantly enhancing interconnection density. It can be seen that ASML is attempting to replicate the technological barriers of lithography machines—precise alignment and high-precision motion control—into back-end equipment, seizing the cake that originally belonged to equipment manufacturers like Besi and Applied Materials. ASML's Chief Technology Officer Marco Pieters previously stated that the company will continue to assess the long-term development trends of the semiconductor industry, focusing on the research and development of equipment bases required for packaging, bonding, and other areas, to prepare for related business layouts. ASML's move sends a clear signal: the war among equipment giants has evolved from a single process to a chain competition across the entire chip manufacturing process. Whoever can provide system-level solutions from front-end to back-end will have greater influence in the next round of industry reshuffling. Canon: Finding "singularity" in a niche In the context of Nikon's collapse and ASML's dominance, Canon has chosen a third path. As one of the three giants, Canon also missed the EUV era. However, unlike Nikon, which confronted the high-end market head-on, Canon is well aware that it cannot catch up with ASML in the wavelength race and has pragmatically turned to focus on differentiated survival. On one hand, Canon is deeply engaged in the mature process lithography machine market. Relying on its accumulation in the optical field, Canon offers cost-effective products, firmly maintaining its base in the mature process markets such as i-line and KrF, which has garnered high loyalty among second- and third-tier wafer fabs. Although its technology level is lower than ASML's EUV and ArFi equipment, Canon thrives in niche markets, flourishing in mature process fields such as power devices, sensors, display drivers, and advanced packaging. On the other hand, Canon is embarking on a new exploration: Nanoimprint Lithography (NIL). This technology operates on principles completely different from optical lithography. NIL does not use complex optical systems to project patterns onto wafers; instead, it directly stamps a template with circuit patterns onto the photoresist on the wafer, similar to stamping, and then cures it with ultraviolet light. Theoretically, the advantages of nanoimprinting are significant: its resolution can rival or even exceed that of EUV, with costs only one-tenth of EUV systems, and the single-wafer processing cost is about one-fourth of EUV; energy consumption is reduced by more than 90%—the total power of EUV systems can reach 1 megawatt, while NIL only requires about 100 kilowatts. In 2014, Canon acquired the nanoimprint company Molecular Imprints Inc. and launched its own technology brand J-FIL. In October 2023, Canon officially launched the FPA-1200NZ2C nanoimprint lithography system, claiming it can be used to produce 5nm chips, with the potential to even reach 2nm in the future. SK Hynix has already introduced nanoimprint equipment from Canon, planning to use it for mass production of 3D NAND flash memory. This is a complete bypass of the EUV system; if nanoimprinting can achieve large-scale production in the storage chip field, which has a higher tolerance for defect rates, Canon may directly rewrite the rules of the game. However, the commercial path for nanoimprinting is still fraught with challenges, with template lifespan and defect control being the two core issues facing this technology. Due to the direct contact of the template with the wafer, the nanoscale structures on it are extremely fragile. Current mass production tests show that the template lifespan can only support imprinting about 50 wafers, far less than the 100,000 wafers lifespan of optical masks. Canon claims that the new design can extend this lifespan tenfold, but industry tests remain unsatisfactory The more fatal issue is the defect replication problem: any tiny defect on the template will be replicated on all wafers, causing serious repeated defects. To detect template defects, the required equipment capacity is equivalent to the global supply of mask inspection equipment for an entire year, which is clearly not economically viable. In addition, the overlay accuracy and capacity of NIL still lag behind ASML's EUV system. Due to Canon's use of a single-wafer platform architecture, it cannot perform measurement and imprinting simultaneously, with a maximum capacity of only about 25 wafers per hour. As the industry describes: “NIL is like a perfectly designed precision clock, far superior in performance and cost to competitors, but the key gears are made of glass—seemingly perfect, yet unable to withstand actual operation.” Canon is clearly aware of this and continues to invest in research and development. In January 2026, Canon announced the world's first development and practical application of a breakthrough wafer flattening technology called IAP (Inkjet Adaptive Planarization), which utilizes accumulated nanoimprint technology to control the topography fluctuations on a 300mm wafer surface within 5nm, with plans for commercial use in 2027. This can be seen as a derivative application of nanoimprint technology, bypassing core challenges and seeking breakthroughs in niche areas first. Canon's path provides the industry with a revelation: When the mainstream technology route has been monopolized by giants, latecomers do not necessarily need to confront them head-on. Finding technological singularities in the gaps and building differentiated competitiveness around long-tail customers can also win survival space. Review and Insights: The Rules of the Lithography Machine Battlefield Have Changed Reviewing the development paths of the three giants in the global lithography machine market, it is not difficult to find that the current lithography machine market has formed a pattern of “ASML reigning supreme, Canon holding a position, and Nikon falling behind.” The fortunes of these three companies reflect profound changes in the global lithography machine industry, releasing many thought-provoking signals and providing valuable insights for other companies in the industry. The Game of Corporate Genes The dilemmas faced by Nikon and Canon largely reflect a common issue in Japanese manufacturing when facing disruptive technological changes—path dependence and perfectionism. Japanese companies often pursue extreme perfection in technology research and development, and once they invest resources in a certain technological route, it is difficult to easily abandon it. This "path dependence" makes it challenging for them to quickly adjust their strategies in the face of new technological waves, ultimately missing opportunities. Especially in the face of disruptive technologies, past successful experiences often become the biggest burden. Nikon's hesitation in the face of immersion technology is essentially a confidence in its "vertical integration" model—self-research of all core components to ensure absolute quality. However, as the complexity of lithography machines increases exponentially, this closed system becomes an obstacle to innovation. No single company can master all cutting-edge technologies alone. ASML's success is precisely due to its openness and collaboration. ASML keenly captured the trend of technological change, quickly adjusted its technological route, and engaged in deep cooperation with top global suppliers, building a vast industrial chain ecosystem. This open collaborative model allows ASML to focus on the integration and optimization of core technologies while leveraging global resources to rapidly enhance product performance, reduce research and development costs, and ultimately achieve industry monopoly The advantages of this open collaboration-based ecosystem may be harder to replicate than going it alone. Behind this is the difference in corporate DNA. The "vertical integration" DNA of Japanese companies emphasizes self-sufficiency and striving for perfection, which can play to their strengths in an era of relatively stable technology; however, in today's fast-evolving and increasingly complex technological landscape, this DNA has become an obstacle to innovation. ASML's open collaboration DNA emphasizes resource integration and adaptability, making it more suited to the industry development trends of the new era. Moreover, ASML's success is also attributed to its continuous innovation DNA. After monopolizing the high-end lithography machine market, ASML did not become complacent but instead keenly captured the opportunity in advanced packaging technology, accelerating its cross-industry layout and transforming from a single lithography equipment supplier to a full-process semiconductor equipment supplier, continuously widening its moat. However, after missing opportunities, Canon has shown greater strategic flexibility compared to Nikon. It did not cling to traditional lithography technology but instead focused on differentiated choices, exploring new paths, and developing long-tail customers, thus gaining its own survival space in the cracks. This wisdom of "knowing when to advance and retreat" is worth learning from for other non-leading manufacturers. Competition Dimension Upgrade: From "Single Machine" to "Ecosystem" The war of lithography machines may have temporarily ended, but the war for semiconductor equipment has just begun. ASML's expansion into advanced packaging marks an upgrade in the competition dimension among equipment giants. With the outcome of the lithography machine market already determined, ASML has begun to leverage its technological barriers in precision alignment and high-precision motion control to extend into backend equipment, attempting to build a full-process solution from "front-end manufacturing" to "backend packaging." According to Yole Group's forecast, the global advanced packaging market size is expected to grow from USD 38-46 billion in 2024 to USD 79-80 billion by 2030, with a compound annual growth rate of 9.4%-9.5%. This incremental market will become a new battleground for equipment giants. This also means that in the future semiconductor equipment market, competition will no longer be about single-point breakthroughs but rather about system-level technological integration capabilities, transitioning from individual links to full-process layouts. Whoever can provide a more complete solution to help customers reduce system complexity and shorten time-to-market will take the initiative in the next round of competition. Additionally, it is worth noting that geopolitical factors will also profoundly impact the future landscape of the semiconductor equipment market. In recent years, the geopolitical game in the global semiconductor industry has intensified, with measures such as export controls and technology blockades not only affecting the development of companies but also altering the industry's supply chain structure. In the future, companies must fully consider geopolitical factors when formulating strategies, building diversified supply chains to reduce operational risks. In Conclusion Nikon's downfall serves as a wake-up call, reminding all technology companies: in this brutal industry driven by both capital and technology, there are no eternal kings, only adapters to the times. Nikon is not lacking in technology or funding. It lost due to misjudgment of new trends, the inertia of a closed system, and the sluggishness in timely adjusting its customer structure. **When the technological route shifts, former assets can instantly turn into liabilities ** ASML's present is rooted in the bold embrace of immersion technology twenty years ago and the strategic vision of building a globally open ecosystem. But history has proven that hegemony is often a prelude to decline. As ASML expands from a leader in lithography machines to a "full industry chain integrator," it also faces new risks: further increases in technological complexity, geopolitical uncertainties, and the threat of potentially disruptive technologies. According to the China Business Industry Research Institute, the global lithography machine market is expected to reach USD 39.2 billion by 2026. In this vast and rapidly expanding arena, the rules of the game have already been rewritten, with shifts in technological paradigms, innovations in business models, and the dynamics of ecosystems that could disrupt existing patterns at any moment. The only certainty is that competition in the semiconductor industry will never cease. Only companies that maintain openness, embrace change, and hold a deep respect for the shifts of the times can survive the next wave of technological upheaval. Nikon's downfall is an elegy of the old era; while ASML's expansion and Canon's exploration are the prologue to a new battlefield. The story of lithography machines is far from over; it has merely turned to a more complex and brutal page