Just for fun, reading some InfoWorld articles from 2004-2005 brings perspective on what the market of electronics hopes for and what it actually manages to release.
Some of those include:
- Intel using Indium Antimonide transistors in the 10s
- Intel using something else rather than silicon after 2013
- IBM releasing Carbon Nanotube transistors in the 10s
- IBM releasing Graphene chips at 250GHz in the 10s
- IBM trying to shove silicon-germanium on AMD for their 65nm process
- Nvidia head scientist expects 10nm and 20TFLOPS by 2015
- Epic Games counting on many, many core CPUs in the 10s
So after 4-5 years on 28nm and even Intel, with their absolutely massive R&D, delaying their chips and even foregoing their tick-tock philosophy for a third refresh on the same manufacturing process, I don't think there's particularly anything that grants optimism. Prototypes have always been done. Like 6nm transistors in 2002 or 22nm SRAM cells in 2006 by the very same IBM...
