Intel is coming this year with the new generation of CPUs on its 14nm manufacturing process. The company announced that it will not use the newer 10nm process in its entire lineup, instead of moving to a better model depending on product segments.Intel will be rolling out 10nm CPUs later this year, but they will be more focused on heat dissipation and slim fanless tablets and ultrabooks. Desktop CPUs will be built on the 14nm strategies.During the presentation, it was understandable that the company is planning to ship 14nm, 8th-generation Core chips during the second half of 2017.
Maybe the most interesting part is how Coffe Lake and Cannon Lake are positioned. Intel is believed to targeting Cannon Lake at lost power ultrabooks, leaving established 14nm Coffe Lake for notebooks and desktops until the 10nm Cannon Lake is revealed.This strategy helps Intel to focus on high-margin products and grow their profits by 30 percent.
Generally, Intel uses two ways to increase chip performance every new release, by focusing on design and manufacturing.This is called "tick-tock" process; new change to chip's design (tick) is followed by a "process shrink" (tock). Both of these methods generate performance improvements. The number of transistors in a given area is doubled every 12 to 18 months as a new boost.Since Intel is leaving its manufacturing process unchanged, does that mean that "boost improvements" will disappear?
Rumors say that Intel is focusing on data center parts "first for next process node", along with ultra-low-power consumer chips.
Intel expects 8th generation to be 15 percent faster than Kaby Lake in some situations.