TSMC started producing 28nm low-power a good year before HP (2010 v 2011). Continuing that rough cadence, and 20nm coming next year, 2014 makes sense for an early form of 14nm (or half-node of 20nm from tsmc). Obviously GF knows what they need to do to remain competitive, and this sounds about right.

Given their track record...this is impressive or just plain optimistic. Either they had some massive breakthroughs and/or got away from impediments by going to gate-last, or this is simply salesmanship they are hoping we forget next time they mention a date and it has changed. We all know fab companies like to do that.

I would love to see both the 28nm chips they produced in 2011, as well as their 20nm chips coming next year...because I think we all know that's for all intents and purposes bullshit...unless they are talking about trial wafers (that have no practical use or meaning to anybody if mass production doesn't follow soon after) or are using tsmc's timeline.

At the rate they have been going I'd expect mass production 22nm at 2014, so somehow I think they are being a tad over optimistic.

Where's the limit of silicon?

Wanna know when Intel has 14 nm tapeout? December...this year.