The acquisition deal emerged out of the close collaboration between the two companies, ASML said (Cymer hs been a supplier to ASML for years). It's likely that ASML determined that EUV development is so critical to the company's future that it simply must have full control over the source development.
But it's also a pretty safe bet that ASML would not have sunk $2.4 billion into Cymer if the company did not believe pretty strongly in Cymer's ability to further improve the source technology. (ASML said it would continue to buy sources from other suppliers, but it only acquired one of them).
Among the more interesting aspects of this deal, ASML will also get Cymer's deep ultraviolet (DUV) source business. ASML says it will continue to provide DUV and EUV sources to its competitors—chiefly Nikon and Cannon. But it's pretty easy to guess who will get top priority for deliveries. The Japanese lithography companies likely would prefer to buy sources from other vendors, if they can.
The painfully slow development of EUV has forced those firms with stakes at the table—Intel, Samsung, TSMC and now ASML—to get in bed with their suppliers like nothing else in recent memory. Although other lithography technologies are still in development and could step up to save the day, these four companies understand that EUV development is critical to keeping the semiconductor industry close to on pace with Moore's Law.
Considering how much ASML's future prospects are already tied to EUV, the decision to double down seems like a good move.
We seem to be transitioning from vertically integrated companies to a whole vertically integrated industry. Too much consolidation will cause problems in the future. Is this cabal of companies already too big to fail? As @resistion says, why not buy all of the other major players essential to the EUV infrastructure while you're at it?
I don't know that ASML has the means or the desire for such a shopping spree. Buying Cymer is one thing; buying every major player essential to the EUV infrastructure? That seems a bit much.
It is interesting, though, that the place ASML choose to buy its way in is the area that is standing in EUV's way. Despite what ASML says about selling sources to other litho vendors, you wonder if they'll keep them all to themselves (assuming of course that the technology progresses to the point where throughput is suitable).
The EUV source is just todays most obvious barrier to success. Tomorrow it will be the reflective optics including the masks, mask inspection and cleaning, and then resists. The companies involved in all of those areas need R & D investment help as well. There are still many obstacles to high volume manufacturing with EUV. And ASML alone wouldn't have to buy all those companies. Intel, Samsung, and TSMC could help. At this point they should just spin off a new company called EUV Inc. and invite a few more big semi companies to join. But that might make it too easy to figure out how much money they are spending on EUV.
Just as the foundry business consolidated as the costs skyrocketed, now the equipment business will do the same. And just as everyone fears the growth of a single foundry domination to be bad for the industry, they will say the same about this segment. The costs just make any other option impractical, like it or not.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.