It's always the same analysts, same vendor, same company people cheering EUV. So I do think it is hype.
But just look objectively at the numbers for 3100 and 3300. It is already admitted that the throughput for 3100 is not sufficient for production. So clearly buying it is simply for learning, indicating that it is largely a mystery with lots of risk for whoever bought it. And that said, the 3300 (designed but not yet built) only offers modest 22 nm resolution, which is not the optical limit but calculated from some other considerations (resist, flare).
I think too much money has been spent money over too long a duration. EUV qualifies as pork barrel. The fact that 193 nm has been extended this far indicates how much EUV has been a let down. In the 1990's there was some work on using STM for lithography. I think 20 years of development could have been targeted on developing the parallelism for a high-throughput approach there.
Blog Doing Math in FPGAs Tom Burke 15 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...