On April 1, 1984, 47 Philips employees were selected and reluctantly transferred to the new joint venture subsidiary ASML. They would work and develop lithography machines in several small simple rooms near the garbage dump. This depressed start-up team may not have imagined that more than 30 years later, ASML would become the world’s largest lithography machine manufacturer, with a market value far exceeding that of its original parent company Philips.
The Netherlands’ largest exporter, the Netherlands’ largest technology employer and the world’s largest chip equipment manufacturer later moved to Wildhoven, a city in the south of the Netherlands with a population of less than 300,000, and continued to influence the development speed of the entire IT industry here. .
Has the lithography machine affected the development progress of the entire IT industry?
To say that lithography technology has affected the development speed of the entire IT industry is not pretentious.
Going back to the starting point, the entire digital world is essentially countless 0s and 1s. For example, every APP, every photo and every short video on the iPhone, all of which are ultimately composed of massive strings of 1s and 0s. These numbers all pass through a chip, which is a network of millions or even billions of transistors. Each transistor is an electronic switch that processes and stores these numbers by turning the current on (1) or off (0). Two numbers. Using your mobile phone to order takeaways, post on Moments, and play games is essentially the chip in the mobile phone, as well as the chip in the Internet platform server, processing countless 0s and 1s at the same time.
The computing power of a computer depends on its ability to process 0s and 1s through the large number of “switches” inside it.
The most advanced computer in 1945 was the “ENIAC” built by the University of Pennsylvania for the U.S. Army. It had 18,000 vacuum tubes as “switches” and was used to calculate the trajectory of artillery shells. It could calculate hundreds of multiplications and volumes per second. It was huge and took up the entire room. Since then, scientists have found smaller, faster, and cheaper “switches”-transistors. By July 1969, the computer that took Apollo 11 to the moon used a Fairchild chip and took up about a cubic foot of space, one-thousandth the size of ENIAC.
In 1961, Fairchild released its first chip with just four transistors embedded in it, but soon the company devised ways to place a dozen transistors on the chip, and then 100… Fairchild United Founder Gordon Moore discovered in 1965 that the number of components that could fit on each chip was doubling every year as engineers learned to make smaller and smaller transistors. This prediction of the exponential growth of chip computing power is the famous “Moore’s Law”. From this, Moore predicted “future products” that seemed crazy in 1965, such as “electronic watches”, “home computers”, and even “personal computers”. Portable Communication Equipment”.
Moore’s Law has almost become the road map for computer development for the next half century or so. By 2020, the A14 processor chip on each iPhone 12 has 11.8 billion tiny transistors integrated into it. The computing power of mobile phones that everyone can buy has far exceeded that of the US Army’s “Eniak”.
More powerful computing power requires lower computing costs. The key lies in smaller “switches” (transistors) and chips with more integrated transistors. This is the charm of Moore’s Law.
The key to maintaining Moore’s Law lies in the lithography machine.
Among the hundreds of processes in chip manufacturing, photolithography is the most important step. A chip requires twenty or thirty times of photolithography during the entire production process, which takes up half of the production process and one-third of the cost.
It is not an exaggeration to say that lithography machines have affected the computing power and information storage capacity of the world’s computers. Considering the wide application of chips in modern national defense and military affairs – the earliest orders in the U.S. chip industry came from NASA and the U.S. Air Force, which were used to guide Rockets and missiles—the pursuit of advanced lithography technology even goes beyond the scope of industrial chain security and has become a focus of great power competition and geopolitical fluctuations.
But when ASML’s first CEO Jared Smit took office, the company was far from the spotlight and could even be said to be a mess.
Before joining ASML, Smit worked as a sales manager for the Dutch office of telecommunications giant International Telephone and Telegraph Company (ITT). He found that profits from the company’s telecommunications business were clearly spiraling downward and were about to bottom out. But after he accepted the company’s CEO position, colleagues at ITT first questioned his career plans, and he also heard from some prominent analysts that the ASM-Philips joint venture was destined to fail.
This statement is not groundless. To a certain extent, ASML is a joint venture subsidiary established by Philips to get rid of its lithography machine project that has been burning money. However, the joint venture partner ASM is not Philips’ first choice, but “make do”. The last straw.
A lithography machine that even Philips can’t afford to “burn”
The working principle of a photolithography machine, or the basic principles of modern chip production itself, is not difficult to understand. The process roughly includes: (1) Drawing a circuit diagram; (2) Engraving the circuit diagram onto a glass plate to make a mask (Also called photomask); (3) Project the circuit diagram on the mask onto the silicon wafer (wafer) coated with photoresist with strong light, and the part of the photoresist illuminated by the strong light becomes soluble. In this way, the circuit diagram is exposed on the silicon wafer; (4) The circuit diagram on the silicon wafer is repeatedly etched, diffused, deposited and other processes are used to create complex transistors and circuit networks.
The fundamentals of lithography machines and chip manufacturing haven’t changed much since Texas Instruments’ Jay Lathrop invented lithography. However, with the development of manufacturing processes (which can be roughly understood as the density of transistors on a chip), the development and implementation costs of photolithography technology are getting higher and higher.
Take the EUV (extreme ultraviolet light) lithography machine that is currently only manufactured by ASML as an example. Let’s look at the light source first. When Lathrop invented lithography technology, only a simple light bulb was needed, but when it developed to the EUV stage By that time, the complexity of light sources had exploded to incredibly high temperatures—in order to generate enough EUV, a small ball of solder needed to be crushed with a laser. Cymer, which has been acquired by ASML, has been a major player in lithography light sources since the 1980s. The company was founded by two laser experts at the University of California, San Diego. Ximeng engineers found that the best way to do this was to launch a small ball of tin, 30 millionths of a meter in diameter, through the vacuum at about 200 miles per hour. Then the tin ball is irradiated with laser light twice. The first time is to heat it, and the second time is to bombard it into plasma at a high temperature several times the surface temperature of the sun. This process of bombarding tin droplets, repeated 50,000 times per second, produces the amount of EUV needed to make chips.
Looking at the lens, at first Lathrop just turned an ordinary microscope upside down, but at the EUV stage, the process precision of the mirror is extremely high. A mirror with a diameter of 30 cm requires a fluctuation of less than 0.3 nanometers, which is equivalent to To build a railway track from Beijing to Shanghai, the fluctuation is required to be no more than 1 mm. Or in the words of Wernick, the current CEO of ASML: “If the area of the reflector is as large as the entire Germany, the highest protrusion cannot be higher than 1 cm.”
Looking at the machine, the area exposed by the photolithography machine at one time is only as big as a fingernail. A wafer with a diameter of 12 inches needs to be moved hundreds of times to expose it all. The positioning of each movement of today’s lithography machines must be accurate to tens of nanometers, which is equivalent to tens of thousands of times the diameter of a human hair. If two vehicles are traveling in parallel at a speed of 30,000 kilometers per hour, the difference between the two must be less than 0.5 mm to achieve the same accuracy as a photolithography machine.
Time is money in the chip industry, and it is common for the prices of unsalable outdated products to plummet. Therefore, the lithography machine must work continuously 24 hours a day, and the annual downtime should not exceed 3%. It is a huge engineering challenge to make such a precise and complex machine work continuously and stably for a long time.
Even back in 1984, when ASML was first established, the research and development of lithography machines cost a huge amount of money because it was still unable to make a profit. Even the wealthy Philips was preparing to cancel this “non-core business”. If the sale did not work, a joint venture would work. In fact, if Trost, the decision maker of the Philips lithography machine project at the time (who later briefly served as ASML CEO), did not use the hidden reserves that only he controlled, the Philips lithography machine project might have been stopped earlier. Philips’ financial executives tolerated Trost’s privileges because they relied on the performance of Trost’s other projects for the company.
ASML: The “reluctant” marriage of Philips and ASM
Originally, the American lithography machine giant Perkin-Elmer had expressed interest in cooperating with Philips. The company had sold thousands of Micralign lithography machines around the world. Intel launched the famous 8086 processor in June 1978, which was manufactured with Micralign. In the late 1970s, Perkin-Elmer had 90% of the lithography market share and had dealings with almost all the top chip manufacturers: from universities to giants such as IBM, Intel and NEC.
With its huge market share, strong customer base and global machine sales channels, Perkin-Elmer seems ideally suited to help Philips out of trouble in the lithography machine market. However, Philips delayed this excellent opportunity because it failed to make timely decisions and respond.
After negotiations with several other potential partners failed, Philips was left with only one option: cooperation with ASM.
ASM CEO Arthur Del Prado studied at Harvard Business School when he was young. While in the United States, he was impressed by the optimism and ambition of the Silicon Valley computer chip industry. A well-known Dutch newspaper later quoted him as saying: When Del Prado returned to the Netherlands, he had a wafer in one hand and $500 in the other. He named his company Advanced Semiconductor Materials (ASM).
Prado was so successful that ASM became the first Dutch company to be listed on NASDAQ in 1981. In 1978, the company had revenue of $14 million; by 1983, revenue had increased sixfold. During the same period, the chip factories of Philips and Elcoma laid off thousands of employees. In Prado’s view, he has been able to build almost all chip production equipment except photolithography machines. As long as he adds photolithography machines, he can become a one-stop chip equipment supplier.
But Philips was very indifferent to ASM’s enthusiasm. The first reason was that ASM was not large enough. In 1980, ASM’s revenue was only US$37 million. According to Philips’ calculations, the research and development costs for the new generation of stepper lithography machines alone will far exceed US$50 million. Secondly, compared with the advanced technology required for lithography machines, the professional technology required for ASM to manufacture wire bonding machines It is simply not worth mentioning. Philips believes that Prado underestimated the complexity of lithography machines; finally, the sales of lithography machines are different from other chip production equipment. The purchase of other chip production equipment can be decided by the management, while the purchase of lithography machines can be decided by the management. Only the board of directors can make sales decisions, so ASM’s sales channels are not helpful for lithography machines.
But in order to save the lithography machine project, Philips finally took the initiative to contact ASM. The meeting lasted only more than an hour. Excluding the time when Prado left the meeting to discuss with the team, the two parties talked for less than 15 minutes, and ASM decided to cooperate with Philips. The lithography machine business fits Prado’s ambitions. ASM manufactures the machines required for every process in the chip production process, but he has not previously been involved in the most strategic photolithography machine.
In the end, the two parties cooperated to establish a 50:50 joint venture company, namely ASML. ASM invested US$2.1 million, and Philips discounted the 17 sets of lithography machine parts in stock for the lithography machine project at US$1.8 million, plus US$300,000 in cash as investment.
Although Prado and ASM failed to survive the “money-burning machine” period of ASML lithography machines and almost withdrew their capital on the eve of the “money-printing machine” period, with a total investment of 35 million US dollars in vain, he did play a critical role in Taking action at all times prevented Philips’ lithography machine technology from being buried, and also contributed to the birth of ASML.
For Smit, serving as CEO of ASML is somewhat of a “reunion”.
He studied aerospace engineering at Delft University of Technology and later received a NASA scholarship to study at the University of Maryland. While working for the European Space Research Organization (now the European Space Agency), he wrote a 70-page booklet analyzing nonlinear solar wind flows in the Earth’s magnetic field. His research confirmed and explained data measured by NASA’s first satellite at the time.
He later joined Philips in 1969, but he chose to leave after a year because he couldn’t stand the bureaucracy within the company.
After his initial meeting with ASML employees, Smit immediately contacted his former colleagues at Philips. They thought Smit was crazy to take on such a “hot potato.”
Let’s look at the team first. The engineers who were transferred by Philips to the joint venture ASML were in an embarrassing situation – they became a joke in the lithography market, and no one believed they could succeed. ASML employees even viewed the new company as a “leveraged carve-out,” the counterpart to a “leveraged buyout” — breaking up a company due to bankruptcy. They all believed that Philips just wanted to get rid of unnecessary burdens.
Looking at existing products, 16 Philips PAS 2000 lithography machines that are about to be produced have also been handed over to ASML. These lithography machines use hydraulic worktables, which require a power unit that is larger than the machine itself, which creates problems of vibration and noise, as well as the risk of oil contamination. The optical components of the PAS 2000 are from France and are not accurate enough. It is difficult to sell a lithography machine with these flaws.
Finally, looking at the market situation, when ASML was established, GCA, the market leader at the time, had delivered hundreds of photolithography machines, and runner-up Nikon was also rapidly occupying the market. What is ASML’s market share? zero.
But after a long conversation with the team, the fog in Smit’s mind began to dissipate. Many elements of Philips’ lithography technology were still leading at the time. Its alignment system’s technology for accurately superimposing chip patterns was very advanced, and Philips’ Natlab had actually An electric wafer stage was developed to replace the hydraulic stage, which is an advantage that competitors do not have. At that time, Natlab was a legend in the world. Energy-saving light bulbs, portable X-ray machines, rotary shaving systems, and video recorders all came from the laboratory. In order to protect their own interests during the joint venture process, the contract signed by Philips and ASM is very demanding, including that ASML pays Natlab 1.5% of revenue every year as R&D expenses, but ASML also has the opportunity to obtain from Natlab everything it needs to advance the development of lithography machines. technology.
Find money, build machines, sell them
Smit believes that it is the customers who ultimately decide ASML’s fate. At the end of May 1984, he flew to San Mateo, California to attend the SEMICON West exhibition and visited chip manufacturers in Silicon Valley, but he received a blow.
U.S. chipmakers told Smit that installed base (the number of machines running in customer factories) itself is critical. The lithography machine is so complex that it can break down due to very small factors. Chip factories want to keep downtime to a minimum, so after-sales service is crucial. At most times, GCA has hundreds of service engineers working on site. ASML has not sold machines, has no service department, and has no practical experience in chip production, while GCA and Nikon already have hundreds of installed machines.
But the industry pulse at this exhibition gave Smit hope. The chip industry is trying its best to maintain the effectiveness of Moore’s Law, and moving from large-scale integrated circuits (LSI) to very large-scale integrated circuits (VLSI) requires a new generation of lithography machines. . Chip lines will be reduced to less than 1/1000 mm, and photolithography machines will no longer process 4-inch wafers, but 6-inch wafers. This transition will occur over the next two years. The new generation of lithography machines will image 0.7 micron details onto the wafer and enable tighter microelectronics integration. And Smit clearly learned at the exhibition: No one has yet found a photolithography solution for this kind of chip. Machines manufactured by Canon, GCA, Nikon and Perkin-Elmer still use lead screws to move the wafer stage, and their image details cannot achieve the positioning accuracy of less than 1 micron, which is the advantage of ASML technology.
An aviation enthusiast, Smit studied the consolidation of the aviation industry, which saw the number of aircraft manufacturers in the world shrink from 50 to just a few from the time he was in college until he got his PhD. During his tenure at ITT, Schmidt also experienced changes in the telecommunications industry. He knows that new manufacturers have no chance in mature markets unless they can achieve major technological breakthroughs.
Now, Smit and ASML have two options: either close down before opening; or deliver a mature VLSI lithography machine in two years and conquer the market.
Smit uses sports analogies when motivating his team. If you follow his expression, what ASML will do next is similar to the “run and gun” tactics on the basketball court – grab the ball, quickly advance to the opponent’s half, and throw the ball in. ASML needs to find enough players. Money, build the machine in a short time, and finally sell the machine.
Want $100 million?
But at Philips, a new generation of machines may take 10 years to manufacture, but ASML only has two years. Their engineers must break the traditional R&D model. The solution is to split the machine into individual modules, and a professional team will develop each module in parallel. module. The real problem lies in the final assembly stage. The traditional method is to test all subsystems one by one. If there is a problem in one subsystem, other subsystems can only wait. ASML could only shorten the testing and assembly phase from two and a half years to six months by building five prototypes at the same time and five teams working in parallel. At peak times, they needed 250 engineers working at the same time.
The engineering team made an offer of $100 million and believed that Smit would never get the board to accept it. Smit was not alarmed. At ITT, he had handled much larger investments. After rehearsing in his mind for 8 days, Smit submitted this sky-high budget to the board of directors. His speech was passionate, describing the crises and opportunities he saw at the exhibition. He believed that the lithography machine industry would repeat the same process as aviation and According to market rules in the telecommunications field, the R&D investment in new generation equipment is 10 times that of the previous generation equipment, and manufacturers will be eliminated in every round. There are about 10 companies now competing for market share, and only a few will be left.
”Therefore, we must at least be in the top three.” Smit explained, “To be in the top three, there is only one way – to invest to be at the top of the industry. We must strive for gold medals, third place is not enough Yes, we must strive for first place. Our only chance of winning is to develop an aggressive, innovative, and focused strategy. The inevitable shuffle shows the cruelty of the market. If we are lucky, we will eventually reach the top; if we do not Too successful and we’ll end up in third; if we’re unlucky, we’ll finish in sixth. But if we’re happy with third or sixth, then we’d better call it a day Stop it. We have to aim for the top, there is no other option. This is our only chance of survival.”
Smit left the board with a choice: pay up, or let ASML go out of business.
The board of directors’ tentative answer is “yes”, and Philips and ASM have decided to increase their investment by US$1.5 million each. The board wanted Smit to find investment and develop more detailed plans on his own, and now he could push the team to build the machine. They named the new product the PAS 2500 and planned to display the machine at the 1986 SEMICON West show.
Since we are rushing to work, we cannot learn from Philips
Smit infused ASML with an “informal culture” that was completely different from Philips. In order to motivate the initially low-morale team, he would even ask people to make cartoon slideshows, which was completely new to engineers who had previously worked at Philips. thing.
In order to quickly recruit a large number of engineers, ASML put out the first recruitment advertisement. The logos of ASM and Philips were very eye-catching, making people think that working at ASML is working at Philips. This earned Smit a scolding from Philips, but he pretended he didn’t know he shouldn’t. The advertisement also said that there was no need to reply to the letter and send a resume. Those interested in applying can call between 6pm and 10pm. This approach was very special at that time. In fact, ASML started the first round of elimination over the phone. The poor economic situation left many engineers looking for jobs in the Netherlands, and the advertisement attracted about 300 applicants.
ASML’s production method is also completely different from Philips. In order to pursue R&D speed and shipment number, it is impossible to manufacture everything by itself, so outsourcing as much as possible is one of the company’s key strategies. In the first few months of its establishment, ASML determined the company’s positioning: a company that only conducts research and development and assembly. This was unheard of at the time.
Because some of Philips’s subsidiaries could not deliver on time, ASML often had to find other suppliers. In the late 1980s, small suppliers in some niche fields could receive orders of the size of ASML, which was enough to maintain operations. The seeds of ASML’s renowned outsourcing ecosystem were planted during those years.
To manage the vast supply chain and production process, ASML requires developers to be involved early in production, allowing them to select components. Every adjustment, including the adjustment of every screw and nut, must be faithfully recorded in the logistics system. This will not only allow suppliers to be more clear when preparing goods, but also allow ASML to better understand the arrival progress of each component. In order to guide 10,000 parts from the river to the sea, ASML even spent millions of dollars in its early days to purchase logistics and supply chain systems from Xerox and hired a full-time employee to oversee the information entry of the system. .
At this time, the first signs of economic recession have emerged, and chip manufacturers are becoming increasingly cautious. Therefore, Smit hopes that customers can start trialling the machine as soon as possible, rather than waiting for the PAS 2500 in two years.
So the ASML team built the transitional machine PAS 2400 in 6 months. Based on the PAS 2000, the hydraulic machine table was replaced by a Natlab electric machine machine. In this process, ASML engineers who were rushing to work at the pace of a startup company had to struggle with the pace of Philips. If an ASML engineer called on a Friday afternoon asking for parts, Philips would say, “It’s impossible to get the parts out before the weekend because it’s almost 5 o’clock.” Philips employees never work overtime. The solution turned out to be: give them a little cash, a few beers, a bottle or two of wine. ASML engineers always kept their trunks full of beer and wine in order to pick up ordered parts from Philips as quickly as possible. In addition, they also give workers some overtime pay in cash from time to time.
At the 1985 SEMICON West exhibition, an ASML engineer compared the PAS 2400 with the competitor’s machines in the exhibition brochure. The other machines he was familiar with had problems during the demonstration, and the booth was often closed during maintenance, while the PAS 2400 was almost closed. Always running.
In order to catch up on PAS 2500, some ASML employees will work overtime until late at night. ASML rented a house in Wildhofen, where engineers who live far away from home can sleep directly if they work overtime until late at night. In case all the beds were occupied, they put their sleeping bags in the trunk of the car.
The PAS 2500, which was originally scheduled to be completed on January 1, 1986, finally caught up with the SEMICON West exhibition in early May 1986. An ASML engineer observed a competitor’s booth and conducted a brief survey. His first question: Which company has the best lithography machine? The opponents all answered: us. Next question: Who has the second best machine? Opponents all answered: ASML.
How to win over AMD, Micron, and TSMC
In early 1986, ASML welcomed its first customer, MMI, a small chip manufacturer, purchasing the PAS 2400. Since then, ASML has finally truly become a new competitor with an installed base. Although the PAS 2400 was only a transitional machine, MMI was so pleased with the machine that its head of production allowed ASML to use his photo in an advertisement placed in early 1986.
The appearance of the PAS 2500 at the 1986 SEMICON West exhibition also attracted another customer, Cypress, and the company’s CEO Rogers made quite a few requests. “You have the best machines in the world,” Rogers told Schmidt, “but if the machines screw up my project and force me to jump off the building, I want to make sure you jump too. So, you have to buy some of my company’s Shares.”
This kind of community of interests connected by equity relationships is actually not uncommon in the chip industry. For example, in order to strengthen its relationship with key supplier Zeiss, ASML later also took a stake in Zeiss’s Semiconductor Manufacturing Technology Group (SMT).
Later, ASML’s chief financial officer managed to arrange funds for the share purchase with NMB Bank, and ASML also received an order from Cypress.
But the customer Smit wanted most at the time was AMD. For this reason, he did not hesitate to publicly “scream” to AMD CEO Jerry Sanders.
At the SEMICON West spring banquet, Sanders lamented that the quality and service of American chip equipment manufacturers were so poor that they had to buy equipment from Japan. So Smit placed an advertisement in an industry journal with the headline: “We heard you, Jerry.” The text of the advertisement: “ASML lithography machine accepts Jerry Sanders’ reliability challenge and guarantees 90% That’s almost twice as long as the industry has now. Jerry, you don’t even have to worry about an earthquake on the San Andreas Fault, our machines are indestructible.” In the fall of 1986,
AMD asked managers to prepare in advance Good to buy PAS 2500 instruments. But at the last minute, Sanders didn’t place the order. At that time, the industry was still in recession and he had no funds. He wanted to wait until the market recovery became clearer before making a decision. This canceled order accounted for half of ASML’s production capacity at the time.
It was not until June 1, 1987 that AMD signed a contract to purchase 25 PAS 2500 units. It was not Smit who convinced AMD, but ASML’s machine in MMI. When AMD acquired MMI and took inventory, it found Perkin-Elmer’s machine gathering dust in a corner. At the same time, six PAS 2400s were continuously manufacturing wafers with the support of ASML service engineers, which allowed AMD to finally say “agree” to ASML.
The same story repeated itself again in the early 1990s. Samsung proactively contacted ASML after visiting the factory of ASML customer Micron, and the two parties reached a cooperation after twists and turns of negotiations.
In the late 1980s, ASML acquired two key customers.
First up is Micron. After consultation, ASML decided to assign a team of service personnel to it, with the goal of making the PAS 2500 meet all specifications promised by ASML, which is to increase the average daily wafer throughput and reduce the maximum downtime of the machine. ASML also tacked on a condition: If the machine’s performance improves, it will share in the profits. In the years that followed, Micron grew steadily, and ASML benefited. While most American companies have ceded the memory field to the Japanese, Micron still persists in production and is now one of the world’s largest manufacturers of semiconductor storage and imaging products.
The other one is TSMC. When TSMC was established in 1987, Philips received US$58 million in exchange for its chip technology for 27.5% of its shares. TSMC is also considered a subsidiary of Philips. It took full advantage of this advantage and made ASML’s negotiations difficult. TSMC refused to pay for the services and finally sent ASML a two-fist-thick contract.
At the end of 1988, as soon as TSMC completed the machine installation work, it sent a fax: 17 new machines were needed because the factory was burned down. This order provides ASML with breathing space during a critical period. Of the machines that were returned, several had only minor smoke damage, and many were easily repaired. In 1989, TSMC’s insurance company, the financial backer that actually paid, became ASML’s largest customer that year.
But less than four months after ASML received the AMD order, Smit resigned as ASML CEO. From a financial perspective, ASML was in worse shape when Smit left office than when it was founded in 1984. By the end of 1987, the company had spent nearly $50 million. The planned sales target was not achieved and ASML continued to lose money. The honeymoon period for ASM and Philips is over. But ASML already has a creative and self-reliant development team, and its logistics and mass production systems have also matured.
Smit was complained by ASML’s CFO as a “big spender”. Even though the company had been losing money, Smit still used external consultants who charged US$700 a day, and even paid for the consultant’s travel expenses to the United States. This is Of course it will cause dissatisfaction among those around you. But if it weren’t for the fact that he valued opportunities over costs, insisted on investing during the industry recession, and rushed to the top at the beginning, ASML would not have gained the dominant position in the field of lithography machines.
Survive the darkest moment when wages cannot be paid
Looking in the rearview mirror now, from 1984 to 1987, the long-term market recession actually gave ASML breathing space. If there had been no recession, Canon and Nikon would have probably occupied the entire market, because ASML and the American lithography machine giant GCA’s The production capacity of key supplier Zeiss is too poor. Even with large orders, ASML’s production capacity in 1986 and 1987 simply could not be completed. The recession will have a much greater impact on Canon and Nikon than on ASML.
In the spring of 1988, ASML experienced the darkest moment when it was about to be unable to pay wages, and it relied on a transfer of US$1.3 million from Philips to survive. In the same year, ASM withdrew its capital in order to avoid being brought down by ASML, and Philips assumed ASM’s shares and debts in ASML.
What finally turned ASML around was the PAS 5500. This machine implemented a modular system like Lego and could be disassembled and assembled like a model kit. With previous lithography machines, chip manufacturers often had to shut down production for weeks and spend a lot of money when lenses had to be replaced.
ASML prepared a “performance” of assembling the PAS 5500 for potential large customer IBM. When IBM could not fly to the Netherlands to see the machine due to the international situation, ASML chose to record the “performance video” and go to IBM. As a result, this equipment was advanced. The extent makes people at IBM very excited.
In the first few months of 1993, capital flows into ASML began to accelerate. Order volumes and deliveries are both up, the latest lithography machines are selling for much higher prices, and revenue from services and upgrades is also increasing. In 1992, the company’s annual revenue jumped from $81 million to $119 million, and ASML could now finally survive on its own. Although the company still lost $20 million that year, this was mainly due to the “labor pains” before the birth of PAS 5500. PAS 5500 keeps the company’s cash flow growing. For the first time in the company’s history, money is flowing in, not out.
ASML used a check to repay Philips’ $21 million “blood transfusion” in May 1992. One day later, the head of Philips’s financial department called and asked ASML never to pay such a large amount of money by check again. This would cause the company to lose money. Two days’ interest.
In 1995, ASML was successfully listed, but not in the Netherlands. The feedback they received during the road show was very cold, and even pension funds did not believe them. They had no choice but to turn to Nasdaq. Listing in the United States and the introduction of American shareholders may be the reason why ASML was able to avoid some geographical restrictions later.
By 1996, some ASML employees started walking around wearing printed T-shirts that said: We will defeat the Japanese.
Defeat Japanese companies
Immersion lithography is the key node for ASML to defeat Nikon.
In the late 1990s, the lithography light source was stuck at 193nm and could not progress. Moore’s Law was blocked. Scientists and industries proposed various solutions.
What finally won was the simplest solution in engineering, which was to add 1mm of water above the wafer photoresist. Water can refract light wavelength of 193nm into 134nm. Immersion lithography successfully crossed the 157nm mark and directly reached the half-cycle 65nm. Coupled with the subsequent continuous improvement of lenses, photoresists, and FinFET and other technologies, the immersion 193nm lithography machine has been able to achieve the 7nm process (the A12 chip equipped with the iPhone XS uses the 7nm process). Wernick, the current CEO of ASML, once said: “The iPhone appeared because of immersion lithography technology.” In
2002, Dr. Ben-Jian Lin of TSMC proposed an immersion 193nm solution, and ASML developed a prototype within a year. TSMC has also become the first company to achieve immersion mass production, and has since caught up with Intel, which previously led the process.
The immersion improvement is small, the effect is large, and the cost is low. Almost no one orders the 157-nanometer dry lithography machine that Nikon launched almost at the same time. Although it only took Nikon a year to catch up with immersive technology, ASML has already won orders from many large customers such as IBM and Intel.
This resulted in Nikon no longer being strong. Nikon was still the leader in 2000, but by 2009 ASML’s market share reached nearly 70%, far ahead.
The difficulties of EUV lithography machines have been described previously. An EUV lithography machine has more than 100,000 parts, requires 40 containers to be transported, weighs 180 tons, and takes more than a year to install and debug.
As early as 1997, facing the difficulty of challenging 193nm, Intel persuaded the Clinton cabinet, which is the most open-minded to high-tech in the United States, to launch the cooperative organization EUV LLC. The organization is led by Intel and the U.S. Department of Energy, and also includes Motorola and AMD, as well as the three major U.S. national laboratories – Lawrence Livermore Laboratory, Lawrence Berkeley Laboratory and Sandia National Laboratory, investing US$200 million to gather Hundreds of top scientists have theoretically verified the feasibility of EUV lithography.
The U.S. government is still very sensitive to the trade war with Japan in the 1980s and does not want Japanese companies such as Nikon and Canon to cooperate with U.S. national laboratories, although Nikon originally believed that EUV technology would not work.
The result was that Nikon was excluded and ASML was allowed in (after making a bunch of promises to contribute to the United States).
In 2012, ASML asked Intel, Samsung and TSMC to invest in itself because the R&D investment in EUV lithography machines required 1 billion euros per year. In total, ASML successfully raised 5.3 billion euros from the three giants. For the whole of 2012, ASML’s sales were only 4.7 billion euros.
In 2015, a mass-produced EUV lithography machine prototype was released, and ASML stood at the top of lithography technology.
As René Rejimek, the author of “Lithography Giant: The Rise of ASML” said in an interview with “China Economic Weekly”: If you want to build a lithography machine, you need to be technically advanced. A lot of money and manpower are invested, but such technology can only last for a few generations. To replicate ASML’s success, companies need to have capital, talent, government support, and historic opportunities at the same time.
ASML’s success cannot be replicated, but its experience can teach newcomers in the chip industry, especially its unique outsourcing methods and supplier ecology.
Today’s chip industry supply chain is complex. A typical chip may be designed by a team of engineers in California or China using American design software and based on the blueprints of the UK-based ARM company. After the design is completed, it will be sent to a factory in Taiwan, China. That factory will then purchase ultrapure silicon wafers and special gases from Japan, and then use the world’s most sophisticated machine made by a Dutch company to engrave the aforementioned design on the silicon. Without these companies, it would be difficult to make advanced chips. The chips are then packaged and tested, usually in Southeast Asia, before being shipped to China and put into a phone or computer.
Still taking the iPhone as an example, TSMC and Samsung, as foundries, are suppliers to Apple; ASML, as a manufacturer of photolithography machines, is a supplier to TSMC and Samsung; and Zeiss, as an optical component manufacturer, is a supplier to ASML…
According to ASML’s 2022 annual report, the total number of suppliers is approximately 5,000, of which 1,600 are in the Netherlands, 1,300 in North America, and 1,350 in Asia.
This long chain shares R&D costs and commercial benefits.
Some analysts believe that a major reason for the Soviet Union’s lagging behind in chips is the lack of an international supply chain. The global division of labor created by Silicon Valley in cooperation with U.S. allies is extremely efficient. At that time, Japan mainly produced memory chips, and the United States produced more microprocessors. Japan’s Nikon and Canon, and the Netherlands’ ASML divided up the lithography equipment market. Workers in Southeast Asia do most of the final assembly. Companies in the United States, Japan, and Europe compete for position in this chain, but they are all able to spread their R&D costs over a semiconductor market that is far larger than the Soviet Union, and thus benefit.
How to make good use of the international supply chain and how to position oneself in the international supply chain are issues that every latecomer in the chip industry should seriously consider.
This article is synthesized from:
(Dutch) Reni Regimek; translated by Jin Jieban: “Lithography Giant: The Road to the Rise of ASML”, People’s Posts and Telecommunications Press, 2020.10.
(U.S.) Written by Chris Miller; translated by Cai Shujun: “Chip Wars: The Battle for the World’s Most Critical Technology”, Zhejiang People’s Publishing House, 2023.5.
Yu Sheng: “Chip War”, Huazhong University of Science and Technology Press, 2021.11.
”Jin Jie Ban” public account “Battle of Lithography Machines” series of articles composed
by the UP host of Station B “Talks Three Circles” popular science video on lithography machines and chips series.