It seems strange how every time a scientist, physicist or otherwise, says shrinking silicon semiconductors is will “soon” not be possible anymore, he is proven wrong by the determination of the tech companies to achieve or maintain high profits.
Intel is reportedly already working and planning the retrofitting and developing of the 14 nm manufacturing process at the company’s FABs in Oregon, Arizona and Ireland.
The company is currently still working on getting more of its FABs up to the 22 nm standard and, in the next years, the presentation shows that Intel’s Israel FAB28, Oregon FAB D1D/C and Arizona FAB 32/12 are the next sites to be retrofitted for 22 nm production.
Later in 2013, D1X Oregon FAB, Arizona FAB 42 and Ireland’s FAB 24 will enjoy the new equipment for 14 nm production that will likely start only in 2014.
When it comes to the 10 nm, 7nm and 5nm manufacturing technologies, Intel lists those as being “in research” for 2015 and beyond. It is very likely that the first new node beyond 14 nm, will be the 10 nm node and that will only be mass produced beginning with 2016.
One thing is made clear by Intel’s presentation material – 22 nm manufacturing technology will be the company’s most advanced process for this year and the next.
The company obviously needs more 22 nm manufacturing capacity, as it’s currently suffering in the mobile area with a 32 nm Medfield chip that is unable to successfully enter the market this year.
The main reason for Medfield’s lack of success is the rather high battery consumption when compared with ARM’s Cortex A9 and A15 designs.
If Medfield were a 22 nm chip, even if the performance were mid-to-high range, the battery consumption would have probably been the best of the pack and Intel’s would have had a chance.