Shaping Moore’s Law in the semiconductor industry: the success story of ASML
In 1984, the year of ASML’s spin off from Philips, you could fit about 2200 transistors on one square millimeter of a computer processor chip. Today, it’s almost 8.5 million1. ASML produces the machines that made this remarkable development possible.
From zero to a global market leader in 30 years: the business model of a winner
ASML produces lithography machines, used by chip manufacturers like NXP and Intel for the production of chips that, for instance, store data in your laptop2. The design of the lithography machine determines how small the features on a chip can be, and therefore, how powerful the end-product for the consumer will be. In this sense, ASML, being at the beginning of the semiconductor value chain, sets the boundaries for what is technically possible throughout the chain.
In 1965, Gordon Moore, one of the Intel cofounders, predicted that the number of transistors on a chip would be doubling every eighteen months. Ever since, the semiconductor industry has lived up to ‘Moore’s law’ and continuously improved the transistor count on a chip1, significantly contributing to the economy’s productivity growth overall3.
ASML has played a central role in this achievement. Through living by its simple business strategy of producing machines that ‘allow customers to produce chips with the highest performance and lowest product costs’, it has managed to deliver increasingly sophisticated equipment. Its success in doing so is reflected in its current market share of 80%, achieved at unprecedented speed, in the semiconductor industry4.
Executing on innovation
Key to ASML’s success is its execution of the business strategy through a three pillar operational strategy: technology leadership, customer and supplier intimacy, entrepreneurial people5.
First, technological leadership is reflected in ASML’s commitment to high R&D spending. In 2014 R&D spending was EUR 1.1 Billion, or ~19% of its annual sales (compared to an average 3% for automakers4). Given the highly cyclical nature of the semiconductor industry, ASML’s constant level of R&D spending stands out, since R&D budgets are regularly cut in economic downturns. ASML, rather than cutting spending on R&D during economic downturns, found other ways to stay nimble. ASML hires roughly 1000 to 1500 temporary workers in economic upturns (compared with ~6500 permanent workers). This allows for flexible adjustment to the business cycle. In addition, the company holds significant excess cash so that in a downturn, there is no need to significantly cut back on R&D expenses6.
Second, customer and supplier intimacy is central in ASML’s approach to innovation. ASML believes in ‘open innovation’, characterized by collaboration with customers, suppliers and academic researchers. Lithography machines are extremely costly to develop – with average selling price at EUR 70 Million1 – making innovation expensive and risky. By collaborating with third parties, ASML can share the risk and success of innovation – as well as make customer input a central part of the ideation process. In 2012, ASML launched a formal ‘customer co-investment’ program to help accelerate its development of Extreme Ultraviolet (EUV) technology. Intel, TSMC and Samsung each agreed to contribute EUR 1.4 billion8. Furthermore, to ensure close ties with academic research, ASML founded the ‘Advanced Research Centre for Nanolithography’ in Amsterdam7.
Finally, ASML stimulates a culture of entrepreneurship in which people are given a lot of ownership and freedom. Teams are close-knit and organized in a multidisciplinary way, and this together with the culture stimulates the exchange of ideas across all layers of the company9.
Operational model as a key driver of the business success
Innovation has been key in the semiconductor industry, and as a market leader ASML has understood this principle very well. Its business strategy is built on continuing innovation and making customers and suppliers a key part of that effort. In my view, their operational strategy has been a driving factor in their business success. Throughout the economic cycle, ASML prioritizes R&D, and has found creative ways to be able to remain committed to this – and has made their customers an integral part of the effort. Finally, ASML’s innovation edge is supported by a culture of entrepreneurship across each of its multidisciplinary teams.
1: In 1985, the Motorola 68020 had 2.235 transistors per square millimetre of chip. Today, the Intel Xeon Haswell A-5 chip has 8.411.498 transistors per square millimetre. These figures are calculated using data from https://en.wikipedia.org/wiki/Transistor_count.
2: https://www.asml.com/asml/show.do?lang=EN&ctx=51431&dfp_fragment=strategy_1
3: US productivity growth 1995-2000: Understanding the contributions of information technology relative to other factors, MGI, October 2001
4: https://staticwww.asml.com/doclib/investor/misc/asml_20140425_ASML_CO_2014-Apr-25.pdf
5: https://www.asml.com/asml/show.do?lang=EN&ctx=50049&dfp_fragment=technology_1
6: http://managementscope.nl/magazine/artikel/513-bert-savonije-asml
7: https://www.asml.com/asml/show.do?ctx=47672
8: http://www.talentbox.nl/index.php/network/cust/161865bd66091aa8b0bdaf327c8f9b01/ASML
Pictures:
https://www.semiwiki.com/forum/content/2952-semiconductor-market-could-grow-15%-2014-a.html
Great job explaining the linkage between the business and operating models. What I found most interesting is your assessment of the technological leadership, especially with the fact that they hire temporary workers to augment their R&D if needed. That way, they are able to reduce R&D costs throughout the economic cycles, thereby reducing the price they charge their customers. This is definitely an advantage given the cost-competitive sector they operate in. I would be curious to know if companies in other sector can adopt this approach of hiring temporary R&D personnel.
Very interesting points on how ASML structures its operational model to account for the seasonal up and downturns in the market. This allows the company to maintain a standard level of R&D investment, but it may also be that their unique capital intensive structure and dominant market share allows them the flexibility to match the timing of their capital outlays on infrastructure with periods of high profitability. You make the point about the importance of the entrepreneurial culture in the company, so I think it would interesting to see how the company structures its permanent and temporary structures to quickly pass this culture onto a large cohort of new hires.
I think that the future of innovation at ASML likely lies in its ability to balance advances in performance in chip structures while still maintaining the affordability and mass production capability of these chip structures. ASML’s real enemy may be the “Moore’s Law” that you talk about above. In order to continue to sell improved machines and equipment to its suppliers, ASML has to be able to justify that these machines are more powerful and just as cost-efficient as the last set that ASML sold them. As we approach the phsycial frontier of how small quantum mechanics allows transistors to be on chips, Moore’s law will ultimately collapse and ASML must prepare now for the changes and opportunities that this will present to its current business model.
Great write-up, I found it particularly informative to learn about the supply-chain of the semiconductor industry. My question is, as ASML is at the very beginning of the superconductor value chain, how do they manage the raw materials that go into their products, especially considering that the physical properties of the raw materials has a direct impact on the technical limits of the end products? I assume that the chip manufacturers ultimately manage the raw materials, but is there collaboration between manufacturers and ASML on that front?