Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Road Walker who wrote (140734)	8/2/2001 7:16:04 PM
From: Saturn V	Read Replies (1) \| Respond to of 186894

Ref < Is SOI all hype ? > SOI is an old idea since the mid-60s, when a thin silicon film was deposited on a sapphire wafer, and the wafer was processed with a CMOS process. The quality of the deposited silicon film was not as good as a regular bulk silicon wafer, and the yields were disastrous. In addition the SOS transistors had a floating substrate, which gave rise to several hard to predict parasitic electrical effects. The promise of SOS was that it reduced the capacitance of junctions and improved the speed of the transistor. But the lower yields and the floating substrate problems prevented this from being a commercial success. IBM modified the SOS idea by using a convention silicon wafer as starting material, and implanting a large dose of oxygen atoms 4 to 20 microns below the silicon surface, and forming an insulating Silicon Di Oxide layer a few microns below the surface , giving silicon devices with similar characteristics as the the old Silicon on Sapphire. This approach should theoretically have better yields than the old sapphire approach, but any new processing idea has lower yields initially. Todays designs have lower junction capacitance than in the past, so the SOI detractors claim that the gain in performance is not worth the risk of this approach which has a bad history. IBM and others have developed better models for the floating substrate problem, but it is not clear how sucessful such models will be for todays extremely complex designs. If large guardbands are used for the problem, the performance gain shrinks rapidly. Intel and TI seem to think that if the same development effort is spent on a "straight forward shrink " of transistor dimensions, the payoff wil be greater. Even IBM which has pioneered this SOI approach does not appear to embrace it completely. Is SOI all hype ? Perhaps not. What is not clear is whether it will finally work on complex designs, and what will be its performance gains compared to a conventional shrinking of transistor dimensions, and most importantly whether it will have the high yields to make it a commercial success.