According to a blog post by Jean Luc Pelloie, the company's Fellow Director of SOI Technology, 20 nm may represent an inflection point in which it will be necessary to transition from a metal-oxide semiconductor field-effect transistor (MOSFET) to Fin-Shaped Field Effect Transistors (FinFET) or 3D transistors, which Intel refers to as tri-gate designs that are set to debut with the company's 22 nm Ivy Bridge product generation.
Pelloie explains that it is "increasingly difficult to control the vertical electric field between the gate and the substrate while maintaining the channel depletion below threshold and then minimize the leakage current" when shrinking transistors. FinFETs are considered a solution to solve this problem. The engineer wrote that FiNFETs could be used on either bulk or SOI wafers. However, ARM still has work to do and especially investigate the scalability of this technology. "However, 3D devices are clearly on the road for sub-20nm nodes … and FinFET’s time may finally be here," Pelloie wrote.
FinFET history goes back to a December 2000 paper published in IEEE Transactions on Electron Devices entitled "FinFET - A Self-Aligned Double Gate MOSFET Scalable to 20 nm". Intel was first to move to tri-gates and take this technology into mass-production, but was accused of having switched to 3D transistors only because it was not able to scale its SRAMs, which the company traditionally uses to unveil a manufacturing process, from 32 nm to 22 nm. Intel's tri-gate technology was announced back in May of this year. Intel hopes that tri-gate technology will keep up with Moore's Law and especially drive low-voltage and low-power features.
We are all quick to dismiss this CPU and that storage device as too slow, too expensive and too whatever. But we have come a long way from the ZX-81 and C64 in a relatively short period of time.
So, given a 200 year time span, what will we be dealing with, and still dismissing as too slow, expensive and what have you.
I wish time travel was more affordable so I could have a look :-)
It might happen in the next 50 years or less.
This doesn't mean we won't produce chips any more,
but that their quality will be more or less the same:
certainly not twice smaller and faster every 3 years.
Mass production of chips based on a completely different techno are still science fiction today.
@jujuvivi
I think we will hit the wall in 20-25 years before moving on so some other type of computing technology (they are talking about hitting the 8-10nm wall within 10 years). After that it will become about 3D chip design (stacked/layered chips), better design, more efficient software, and moving on to quantum computing or some other technology yet to be invented.
What I think is really crazy/wierd about technology is that my son will likely never have a traditional desktop of his own (or at least not need to), and that kids who are currently computer illiterate will never need to become literate because everything will move to personal sized devices using server/cloud services where all that matters is bandwidth and screen size/quality.
The future is wierd, but it is neat to see where ARM and x86 take us.
Crap, what did you hear?? Are we going to elect another brain dead pot fiend Republican as President... Again...
We've already had crackhead and snake in the grass/false hope prezies. I think pot is the least of our worries.