Exactly right. The nm figure has been misunderstood by the entire investing community. There is no guarantee that a smaller die size is reflected in a more efficient chip. 20nm (KnC Neptune) in my view is a step too far and it's cost of production will not be justified by it's meagre if any improvements over top-spec 28nm.

Again, there is nothing more in friedcats release than in Ken's from last week (plus his posts) except I/O standard (as investors do we need to know that!) and a price point that has a variability of 100%.

Now some people may think a theorhetical chip that hasn't been taped out with a cost per chip prediction that ranges by 100% from a proposer who does not have the funds available to produce that chip is hard news but I would wonder at their sanity.

When he's selling the chip in bulk, selling 10TH/s mining rigs (55nm chip only) with it and running a substantial farm with his chip maybe we can pay him due respect as a competitor. Till  then we just get on with our own projects.