Re: EUV vs XRL according to Dvorak
Full link:
www8.zdnet.com
First Part of Article:
I recall that about a decade ago all the makers of semiconductors talked about the 1-micron barrier as if it were the sound barrier in 1940. Now we're moving toward the 0.25-micron process. Both AMD and Intel expect their 300-MHz microprocessors, which now use the 0.35-micron process, to use the 0.25-micron process probably before the end of the year. When AMD rolled out its K6, it had a bunch of 0.25-micron chips ready to show. I suspect Intel is eager to do anything to reduce chip size.
These developments do not take place in a vacuum or overnight. Years of research and development go into each change. Currently all the research is divided into two main camps, those developing the 193-nanometer ultraviolet laser to be used for the 0.18-micron process, and the newer X-ray lithography (XRL) crowd. The latter group knows that the future is in XRL but also that the industry is going to be reluctant to move to it quickly, since it will require a complete redesign of almost everything and since, the XRL group believes, the changeover will be much more expensive than continuing to shrink spectrum light technologies. The problem with shrinking the old technology is that it seems to be at a dead end at 0.18 or maybe 0.15 microns. Worse, it's possible that even 0.18 microns may not be achievable. There is no lens material that can withstand the million-bursts-per-minute laser pulse without being damaged. The XRL crowd will be pleased if these failures continue, because they may quicken the decision to move to XRL (which begins its process life at 0.13 microns).
Curiously, Intel is funding another research effort called Extreme Ultraviolet lithography (EUV), the goal of which is to achieve 0.13 microns without using X-rays. The key here is a 13-nm beam that requires even more special optics. Insiders believe that if a 13-nm EUV system is developed, then all bets are off.
If XRL happens sooner rather than later (assuming all the efforts to stay within spectrum light are failures), I believe we'll see a momentary industry slowdown as production snafus and price hikes confuse the market. This should come to a head in the next 12 months.
I found all this research interesting in light of a recent survey in R&D Magazine that revealed that more than 80 percent of respondents think America will lose its competitive industry edge within 25 years unless basic-research funding is increased.
- Jeff |
