Quote from: MrTeal on October 09, 2014, 08:11:40 PM

What are you going on about? All you're describing is that some larger groups can get equipment cheaper than others, which is no different than some people getting cheaper power than others. Change $330 per TH/s to something else like $200/TH/s, and you're done. All that does is change one variable. In the end, you still end up with the same result; even the manufacturer won't make more equipment as it is not cost effective to do so, and the hashrate will stabilize.

My original point was that people buying miners are likely to make very little profit if any at all. Implicit in that statement is that this refers to consumers buying these products at "retail" prices. That is the statement I was defending. Which, I agree, is at this point only tangentially related to the discussion that is now going on.

I do agree that manufacturers will not make equipment when it is not cost effective to do so. That much is obvious. But the graph still does not accurately describe this, as, again, implicit in the graph is a technologically static environment, which does not correspond to reality.

As for the discussions about nearing the limits of Moore's Law, or how that corresponds to power efficiency, etc... the thing that people often forget is that the current paradigm of silicon integrated circuits is just one embodiment of the general trend of exponentially accelerating technological progress. There are many aspects of progress that we are just getting started on now: three-dimensional as opposed to two-dimensional elements and ICs, materials other than silicon with better performance, optical computing, quantum computing, etc. As any particular paradigm like just shrinking integrated circuit process size runs out of steam, other ones will pick up the slack. We're not gonna suddenly hit a brick wall where computational power just stands still... at least, not for a very long time, until we start to reach the fundamental physical limits that come from information theory.