Intel’s 10nm process was designed to be ambitious and allow Intel to reduce footprint by over 50% compared to its current 14nm manufacturing node. The 10nm were originally planned for market introduction in 2015, but the process was subject to problems throughout its development, so products with this node are not expected to be on the market until the end of 2019.
The 10 nm node is a headache for Intel
Intel traditionally started two architectures at each process node, creating the popular “Tick Tock” start cycle. The 14nm broke this cycle, with Broadwell, Skylake, Kaby Lake and Coffee Lake releasing the process. Coffee Lake-S processors are now expected to be included in this 14nm product list, demonstrating how difficult it is for Intel to make the leap to a more advanced manufacturing process with fewer nanometers.
The only processors Intel manufactured at 10 nm were exclusive Cannon Lake processors for a small number of Chinese laptops.
According to Semi Accurate sources, Intel is downgrading its 10nm process to get production ready as soon as possible, sacrificing some of its space/area savings in the array. The goal is that they can improve their production performance to such an extent that they make 10nm financially profitable. In fact, Semi Accurate goes so far as to say that Intel’s new 10nm process will be “effectively a 12nm process”, although this is something Intel would never publicly admit.
Intel seems to have been too ambitious with the 10nm and has created a manufacturing node that has proved too difficult and has forced the company to reduce its space-saving goals to make the process suitable for the manufacture of large processors.
Regardless of how you look at it, the 10 nm is one of Intel’s biggest failures in a long time, which will lead to market share losses in the server area next year, where AMD will have its EPYC chips with a 7 nm node ready.
Source
Overclocking3D