SAAA
Senior member
- May 14, 2014
- 541
- 126
- 116
Don't forget 14nm was a 1.5 node jump from 22nm. (16nm would have been a single node jump from 22nm)
With that mentioned, it is true than 14nm to 10nm is only a single node jump.
That's true too, the factor used to be 1.6-1.7x denser than the previous node but this time we got >2x.
Going by the graphs Intel released months ago 10nm should also be close to 2x scaling and cumulative that's a >4x factor. It's like three node jump in two so a 2.5years timeframe isn't bad either.
The largest problem will be getting those node to yield and more economic than 22nm... I can't see that happen easily with triple-quadruple patterning litho.