The implication of the IC Insights analysis is that ....in aggregate for the industry working the trailing edge/specialist edge. or shall we call it More-than-Moore, as worked by Tower Semiconductor, X-Fab, Lfoundry, Silterra and others, will NOT work, because it is a declining market?
Obviously it can work for thoise gaining market shares while others drop out of the market.
But why is it declining in aggregate?
Well it may be that TSMC is dropping ASPs for More-than-Moore wafers as it grabs market share.
And there may be some protection if a specialist has a process that is in demand and not offered anywhere else. But again the implication is that eventually TSMC or Globalfoundries will offer it while expanding their business at the leading-edge.
There are, of course, long time constants on a market that rarely reaches dynamic equilibrium.
I don't see any surprises in the story's headline: Pure-Play Foundries Need the Leading Edge.
The question I would ask, though, is whether that's the most profitable path for pure-play foundries. When you look at profitabilities of foundries, I think we may come to a different conclusion with specialty foundries like Tower-Jazz are doing quite well.