Intel used to have quite steady tick-tock execution. It usually managed to launch a new processor generation consecutively for more than five years, transitioning to a new node on a regular basis, before it slipped with 14nm Broadwell.
Broadwell launched for the most expensive razor ultra-thin laptops and detachables and Intel claims as many as 70 design wins shipping this year. The company is planning to launch the Skylake 'tock' processors in 2015 and this will be its second generation 14nm big core processor. In Intel’s vocabulary, 'tock' is a cute name for the second generation processor based on the same basic architecture and process.
Broadwell is the first iteration of Intel's 14nm manufacturing line and it's a 'tick', so it will be followed by Skylake, a 'tock' in 2015. A year later, at some point in 2016, Intel should be launching its first 10nm processor codenamed Cannonlake. Intel was obviously late with the transition to 14nm, as this manufacturing node was not as easy as they expected.
However, the competition is doing worse than Intel, as AMD is currently stuck at 28nm with hopes of transitioning to 20nm in 2015. Apple is producing millions of A8 mobile SoCs based on the 20nm manufacturing process and Qualcomm is getting ready to roll out and ship its Snapdragon 810, its first 20nm processor.
Intel told investors that the 14nm node will be replaced by 10nm, but it didn’t talk about the actual time of deployment. It even lined up the 7nm node to come after 10nm. If all goes as planned we might see 10nm Intel processors in 2016 and 7nm based Intel processors in 2018, but these are best case scenario timelines, so there is a chance that these dates will be slightly pushed back.
Samsung and GlobalFoundries are supposed to have their 14nm node ready in 2015, while TSMC is doing risk production of 16nm FinFET (16FF), right now with some serious products probably coming in the course of 2015. TSMC's 16nm FinFET process will be complemented by 16nm FinFET Plus (16FF+) in early 2016.