There has never been any "commercial" full-scale xray mask production to date, so to say that the vendors have "dropped" xray litho from their roadmap isn't really saying much in my view. As with any other new technology the supporting infrastructure will not develop to the point of producing masks which are economically viable until the technology looks like it is going to be adopted and there is going to be a demand for masks. Although EUV is on Sematech's roadmap, and therefore the various equipment manufacturers are investing in the supporting equipment, what is your view of the progress and resources being devoted to EUV masks. How about the optics required for the proposed EUV systems?
With xray masks, the question I have is whether there are significant technological barriers to making the xray masks, or is it mainly the issue of the mask makers not yet wanting to devote time or capital to manufacture them at the levels needed to supply commercial xray systems if the demand for xray systems otherwise comes about? Maybe this is a chicken and egg kind of thing, but I dont think it is too unusual or unexpected to assume that the mask makers and resist providers will not throw a lot of their own resources at new mask development (as opposed to a government or EUV LLC or other industry consortium funding such development) until a commercially viable xray litho source is demonstrated.
There is no question in my mind that a huge selling effort will have to be done to reverse the momentum of almost the entire silicon semi industry backing the EUV roadmap and convince a meaningful portion to accept a point source xray litho system. But it appears that as an opening move that there may be some acceptance in the GaAs market. Although the throughput levels are not acceptable for silicon, it seems to be a valid approach to "prove" the technology commercially on a lower level with the hope of getting some credibility and acceptance from the silicon industry (assuming there is also a credible plan that shows how the JMAR source will increase its throughput to acceptable levels for Si).
What is fascinating (and also not unexpected) is the battle whcih continues to wage between the few xray supporters (such as Hank Smith at MIT) who challenge EUV's commercial viability and attack EUV's weak points (expensive optics, low power, etc.) and the EUV supporters who attack xray litho's weak points (low power and masks). Although supported by the major semi manufacturers (Intel, AMD, Motorola, etc), is there any real difference between the people doing the work for the EUV LLC (university researchers and government labs - Livermore, Sandia, Berkeley, etc.) and Hank Smith and his friends? Who's to say that the EUV LLC hasn't just done a better job of selling a multibillion research program to the semi manufacturers? They clearly control the message (as politicians say) with their immense PR budget. So how the hell does an average investor sort out which technology is going to be adopted? |
