The SPIE Advanced Lithography Conference was held last week (12-16 February 2012). I will post a number of gleanings in this series.

The market window for EUV is rapidly closing. Intel is rumored to have locked down the processes all the way down to the 14-15 nm node. This means that the next available node for EUV insertion is 10-11 nm. If that node is missed as well, EUV at the current wavelength is a one generation technology. Since flare scales with the inverse square of wavelength, any wavelength reduction forces the surface roughness of the materials used for making mirrors and masks to far less than 1 nm 3σ for the flare to be held below 10%. I conclude that EUV may never be production worthy without an immediate step-function advance in source power.