Sort-of. The successor (Ice Lake) is planned to be on a 10nm+ process (so an advanced, 2nd gen version of the 10nm process). However, if the issues with Cannon

Sort-of. The successor (Ice Lake) is planned to be on a 10nm+ process (so an advanced, 2nd gen version of the 10nm process). However, if the issues with Cannon Lake are mostly about moving to a smaller feature size (i.e. it's mostly a die shrink), and Ice Lake is more of an arch update than a feature size reach, it makes sense that Cannon Lake gets pushed back more than Ice Lake would be. (Yes, it'd necessarily be later -- but if you tackle most of the hard problems now, maybe not _that much_ later, and it makes sense to skip a gen.)