The race for new methods of manufacturing CPUs is on in earnest. Although most of us cannot see the efforts on the back ground, we do hear little bits from time to time about this or that new technology that could be used in CPU manufacturing.
The problem is that no matter what the new idea is today, it will take the better part of a decade to realize any real changes in the way we make our processors.
So while we have news that IBM is working with DNA strands to find new ways to make CPUs or that Intel is working on Grpahene (a 2 dimensional sheet of pure carbon) and Carbon Nano Tubes. These changes will not see daylight until 2020 or later.
This means that without a radical change in manufacturing technology soon we will see CPUs hitting a roadblock at around 28nm. This is when many foundries feel the existing methods and materials will reach their limitations. We already know that at 40nm current leakage is so bad that both ATi and nVidia are facing crippling yield issues. 32nm with HKMG is better but still might not work at below 28nm.
It will be interesting to see if one of the leading technologies can be put into play before the current material set runs out.