There will be a very large market for 193 immersion tools for several generations to come. While a few most critical layers will require EUVL, the rest will be patterned by 193 immersion. Nikon's biggest problem (and opportunity)is they are behind ASML in addressing that market. ASML is able to continuously innovate beyond conventional equipment parameters like throughput, lenses and stage accuracy. They invented flexible illuminators thanks to their in-depth understanding of the entire patterning process, most importantly the degrees of freedom offered by masks and RET. Nikon needs to provide expertise in computational lithography to complement their tool offerings. ASML's "holistic solution" approach has been right on. Nikon's market share upside is big. They can acquire computational lithography expertise and technology just like ASML did in 2006.
I think the bigger worry should be what to do for 10 nm devices? If something is 10 nm thin or less, is it worth using? Any metal wire that thin would be highly resistive and any interconnect oxide film that thin would be more prone to dielectric damage. Isn't it time to prepare to ditch the entire silicon wafer infrastructure?
There are ASML staff working on imprint, to my knowledge. So ASML does see a danger in betting all on EUV (or on one technology). But they too can afford skip out or slow down on any NGL (as Nikon and Canon appear to be doing), since double triple or quadruple patterning will entail more purchase of their current bread and butter immersion tools.
The multi-beam approach can also complement for ultra-low volumes or masks. Canon I think also worked on ML2 for a while.
Maybe its ASML future that analysts should be concerned about? Some risk development in any industry is expected, but sometimes taking the lead does not translate into a success. Major investments require major success to not have a major negative effect on the bottom line. Perhaps Canon and Nikon are in a better position to survive.
ASML has made such a hugh investment that there is no turning back now, even for the industry as the industry thinks that there are no other alternatives. My question is why Canon and Nikon should go after EUV? Should not they look into the alternative technology like multi-beam which has shown a lot of promise in the recent past. Let ASML chase EUV and let Canon and Nikon chase multi-beam and winner takes all.
Blog Doing Math in FPGAs Tom Burke 24 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 ...