Reading time: 6 minutes
ASML is using brute force to keep complex EUV developments on track. Its ambition to farm out more work has taken a temporary back seat.
‘ASML wants to farm out even more,’ wrote Bits&Chips five years ago. The company in Veldhoven wanted to take the outsourcing for which it is renowned to the extreme. Even more work would be contracted out to its suppliers: not just manufacturing, but also more development.
Suppliers were expected to take on a pioneering role and develop their subsystems themselves. Creating roadmaps, making investments, taking responsibility. ASML also wanted to drastically shrink its pool of partners. A wave of consolidation was imminent. This much was certain: the supply chain would soon be squawking.
In the spring of 2011, ASML set a high bar at its supplier day – for itself and its entire ecosystem. By 2015, manufacturers had to be vested in their new role: not of supplier, but of partner. All of them following the example set by Cymer and Zeiss, the superstar suppliers who developed their light sources and optical systems fully autonomously – albeit in close conversation with ASML.
As expected, the tier-one suppliers Frencken, Hittech, Norma and NTS swelled in size through acquisitions. NTS in particular was an M&A champion. To top it all off, last year the company even took Norma under its wing. On the positive side, Dutch suppliers are getting better and better at taking work off ASML’s hands and, in doing so, taking the initiative themselves, as several articles in this special issue show.
But the shifting of responsibilities wasn’t a simple task. That became evident when VDL ETG was handed responsibility for the development and production of ASML’s wafer handler. Despite both companies’ commitment to seeing the project through to a successful conclusion, it quickly became clear how hard it is to separate out the R&D on a system as complex as that one. After three intense years of effort, the project still isn’t running smoothly.
Yet the case of the wafer handler is child’s play compared to what happened at Cymer and Zeiss. Over the past five years, the desired move toward greater outsourcing had gone in the other direction at these major OEM suppliers: ASML was pitching in to keep the extremely complex development of EUV lasers and the upcoming, second-generation mirror column on track.
The complexity of the EUV endeavour compelled ASML to reach in with an iron fist to ensure a good end for this Apollo programme of the semiconductor industry. Instead of trusting in its OEM suppliers, the company meddled heavily in the development they were doing, culminating in an acquisition and a participating interest.
It started small in September 2012, with the acquisition of Wijdeven, a manufacturer of the linear motors that are crucial to throughput in ASML’s systems. Double patterning and EUV have sent the cost of lithographic machines skyrocketing. Faster scans, and thus more advanced linear motors, help to compensate for that. Apparently ASML didn’t want to be dependent on a relatively small shop for this make-or-break component.
The six million euros the company paid for the motor manufacturer in Oirschot were peanuts compared to the strategic moves that followed. Barely a month later ASML announced it was acquiring Cymer for 1.95 billion euros. The press release left no doubt as to why: the acquisition would accelerate the development of a high-power EUV source.
For those who still questioned ASML’s commitment to getting EUV across the finish line, there was last month’s announcement that for one billion euros, the company would be acquiring nearly a quarter of Carl Zeiss’s optics subsidiary SMT. The driving force here is to squeeze the absolute most out of EUV in the future. In order to continue scaling to the tiniest details after 2020, a second-generation mirror column will be necessary. In addition to the acquisition sum, ASML will be investing a total of 220 million euros over the next six years into SMT’s R&D efforts, and it’s supporting its supplier with another 540 million for other crucial investments.
EUV has required more than anyone ever expected. ASML now freely admits that – certainly now it’s become clear the EUV project will be a success. But the machinery manufacturer in Veldhoven had to pull out all the stops to get there. For years, it stationed a hefty delegation of engineers at Cymer in San Diego to get the source’s development on track. A similar phenomenon is now occurring at Zeiss: ASML-ians are pitching in on site in Oberkochen to perfect the second-generation mirror column.
Cymer couldn’t keep up with developments. The company excels in physics and fundamental research, but it couldn’t wrangle the engineering of the source into shape. If the American laser supplier had been forced to rely on its own technical and financial resources to develop the EUV source, the source would probably have been much too expensive – if one had ever seen the light of day.
Zeiss SMT is strong in engineering and optics, but the Germans won’t be able to free up or attract the talent in the short term to develop a second-generation EUV column on their own. At any rate, we can conclude from ASML’s recent investment that its faith in that was lacking. The Dutch-German tech pact is a move to be safe rather than sorry, by combining the two companies’ strengths. That means that in the coming years, a sizeable crew from Veldhoven will be stationed in Oberkochen to help with the mechanics and mechatronics.
Putting out fires
The fact that ASML is even able to raise the ante this way can be attributed to its peerless success. Not only are its market cap high and its cash reserves sufficient to take control of the effort through an acquisition and a participating interest; the machinery manufacturer has also become one of the most attractive high-tech employers in the world. Physicists, mechanical engineers and experts from other disciplines around the globe know by now that Veldhoven is the place where they can advance the frontiers of high-tech industry.
That appeal has granted ASML a position of luxury over the last ten years. And now, its accumulation of intelligent grey matter seems like an insurance policy for the future. Even suppliers with one to two hundred engineers can’t suddenly free up twenty employees to resolve a problem that’s just cropped up.
But ASML can – and does. What’s more, these engineers bring ASML’s network with them. It’s well known that Veldhoven calls on its suppliers and engineering firms to get things back on track when the going gets incredibly tough. And when ASML sounds the alarm at critical moments, it’s incredibly tough to say you won’t help put out the fire. So not only ASML’s own engineers, but also many other mechanics and mechatronics experts in the Eindhoven area pitch in to help Zeiss complete its optical column or to help Trumpf complete the mechatronics required to fire lasers at droplets of tin.
And we should also remember that ASML can always fall back on an impressive network of scientists and research institutes, too. The region’s universities and institutes such as FOM have strong ties with the company. TNO and Philips Innovation Services are also always ready to help if things get truly difficult. In that regard, ASML’s influence has vastly strengthened the Dutch ecosystem.
And that will certainly come in handy in the future, because the people we asked for comments while writing this article all say that the violence will only increase: larger investments, greater complexity, shorter time to market and intenser collaboration between OEMs and suppliers.