power is the big obstacle, and it's the one thing I don't see the upcoming mining competition talking about. I don't think people realize what it takes to get 10 TH/s online or 50 TH/s, and forget about 500 TH/s. It's not like just plugging it in to the wall socket.
I remember back in June in the russian thread Bitfury used to make quite a big deal of his chips' power consumption. He used to say that in coming month the power would become more and more important factor, until finally it becomes the most important limiting cost (instead of actual chip or PCB manufacturing costs). He used to brag about his chips being so "green", and they turned out to be efficient indeed, although not as much as he predicted. He was even talking about using ASICs to heat his greenhouse garden during winters in a few years, because they would barely pay for the electricity cost, and maybe not even that.
So at least some of that "upcoming mining competition" does take power consumption very seriously.
yes, some groups talk about the efficiency of their chips, but not how they will power a 50 TH/s or 500 TH/s mining operations within a few months (or less). As we have already discussed in this thread, you can rent datacenter space, but it isn't cheap, and I wonder if anyone is including those costs in their plans (they are certainly not sharing the cost breakdown).
I think it is great that some of the smarter groups are looking into efficiency, but efficiency is only one piece of the puzzle. You still have to have hundreds of kW of power available to your mining operation, if you plan to get it online at some point.
The focus of discussion for most of the IPOs tend to be chip design and/or chip arrival. Getting the chips is a major step, but it's only the first step, and certainly not the most difficult (yet, most stumble at that stage)
