But Intel is going about leading the transition in a rather odd way.
When IBM, Motorola and Infineon led previous wafer size transitions they didn’t find it necessary to buy stakes in the tool -makers. But Intel is taking 15% of ASML
What seems disproportionate is that Intel is spending three times as much on ASML shares as it’s putting into ASML’s R&D: $3 billion on shares; $1 billion on R&D.
Intel is the most vocal proponent of the transition which is variously estimated to cost between $20 billion and $40 billion.
This $1 billion for R&D looks insignificant, while the $3 billion going into buying ASML shares looks irrelevant, unless Intel is looking to buy control of the tools industry.
If it did that, it could keep 450mm technology all to itself.
For 450mm development based on existing litho tools, 1 bn is far from insignificant, it’s in the order of magnitude of the expected total R&D cost.
Basically ASML is saying: “I will only make 450 mm if you either promise to buy the machines once they exist, or pay the R&D upfront”, and Intel choose to pay the R&D.
The shares buy is a bit more surprising. The profits (dividend) from ASML will not do anything significant for the pile of cash Intel already has, so I see it more as a symbolic partnership than as a financially driven decision.
Seems entirely sensible to me. If ASML is going to be a critical part of your business, buy into it. They are doing just the same as they and Apple did with Imagination Technologies.
Alternatively they could put money in and have Samsung stroll by and take a big share in ASML…