This news article was quite interesting and may help diffuse a situation between us. My comments on SVGL were more of a timing issue than of a capability issue. Much like I have stated about the use of x-ray and e-beam being put off for many years as lithographers figure out how to make a better mousetrap, the following may stall the advent of the more advanced lasers by some cost effective manufacturers. This most certainly will give the likes of CYMI and their end users, the benefit of time to go head to head with the Micrascans, IF indeed the SGL equipment will be as dominant as you suggest.
MIT lab makes 0.05-micron gates with phase-shifting, 248-nm lithography - By Semiconductor Business News
Feb 21, 2000 (8:36 AM)
URL: semibiznews.com
SAN JOSE--Numerical Technologies Inc. here today announced fabrication of the world's smallest transistor gate lengths, measuring 50-nanometers, with the company's phase-shifting photomask technology and deep-ultraviolet optical lithography.
The 50-nm (0.05-micron) transistor gates were produced by Massachusetts Institute of Technology's Lincoln Laboratory in Lexington, Mass., using DUV 248-nm lithography equipment, said NumeriTech.
"People have been predicting the end of optical lithography for several years, saying that it can't extend beyond 100 nanometers," said Y.C. (Buno) Pati, president and chief executive officer of NumeriTech. "The MIT Lincoln Laboratory results prove that with prudent use of phase shifting, optical lithography can be extended much further than anyone ever thought was possible."
The experimental devices were produced by the MIT laboratory as part of a program sponsored by the Defense Advanced Research Projects Agency (DARPA). The program is focused on sub-100-nm fully depleted silicon-on-insulator CMOS technology.
According to NumeriTech, this is the first time 248-nm lithography equipment has produced a 50-nm device. Today's 248-nm manufacturing tools and photomasks are just now beginning to produce ICs with feature sizes slightly below 0.18 micron. NumeriTech said experts have previously estimated that next-generation 157-nm lithography would be required to print lines at 70 nm. Some estimates show it taking the chip industry up to 10 years to fully develop 157-nm lithography tools and supporting technologies, said the San Jose company.
On March 2, technical details about the fabrication of 0.05-micron gates will be presented by MIT researchers during a conference sponsored by the International Society for Optical Engineering (SPIE) in Santa Clara, Calif.
This is very R&D but at least it was demonstrated in the lab. Does this mean it becomes production worthy? IT really all depends since the cost of PSM is formidable and prohibitive IMHO. That is, of course, unless PSM reticles come down in price as new reticle making techniques are brought on line to make them more cost effective and cheaper to manufacture.
So much for the death of Optical Litohgraphy at 100nm. and to think it was suppose to die at just above 1.0u a decade or more ago. <GGG>
We must all keep our sense of humor these days since we never know what comes in out of left field. And the way I read the Lamda Physik news release (my impression), they are conceding the present laser technology to the likes of Cymi and Komatsu, putting all of their chickens in the advanced basket.
Finally, not to knock either Lambda Physik of SVGL, I always find it interesting when someone bucks the trend and marches to the beat of a different drummer. CYMI owns the market but SVGL is going with Lambda Physik. The history of the Micrascan has many missteps and this could be a killer, if all does not go well. However, I must admit, I do like their choice of Lens suppliers.<GGG>
We Live in interesting times and as an ex-lithographer, nothing amazes me. From 10 microns down to 0.18u, I have lived through it all.<GGG> I admit to my biases and wish everyone the best in their investments. It just means we see companies in a different light at different times. My decision to look at BTUI more seriously was a shift in my past bias towards anything that was not SVGI (Thermco) or TEL. SO, I can change<GGG>.
Andrew |
