Quote from: bcp19 on September 29, 2013, 01:10:41 PM

Quote from: cpengr on September 29, 2013, 12:59:50 PM

Quote from: joeventura on September 29, 2013, 02:48:30 AM

Your math is different then mine.

Here's my math (screenshot below), which is crazy optimistic in every way, except for the non-overclocked, exactly as-specced hashrate. How does yours differ?

Quote from: joeventura on September 29, 2013, 02:48:30 AM

I am even by December

That's roughly consistent with my screenshot above. And it barely grows past that.

If I actually take delivery, I estimate I might get it in late Oct / early Nov. I'm postulating that since it looks like I won't even break even, that I should request a refund--and I'm here looking for second opinions.

Do you realize that by April you have the network at 120PH, May 240PH, June 480PH, July 960PH? This growth is unsustainable. The biggest gun out there at the moment is 2TH. It'd take 60k of them in April or 2000 a day. 120K of them in May or 4000 a day. 240K of them in June or 8000 a day. See where this is going?

Here's another way of looking at the potential for continued long-term exponential growth (but I think it uses the same line of thinking). In the days of GPU, it took us 300-400 watts per GH/s. The network was at around 23TH/s total at the beginning of the year. So, there were 7-9 GW of energy dedicated to mining and "normal" network growth.

The first generation of devices now can convert 6-10W per GH/s. It took us about 10 months to get to this point, 1300TH/s as people convert their devices over, cash out, etc. So, with 30-70x more efficiency, we should expect 30-70x more hashrate, let's call it 1610TH/s. We're nearing the end of round 1, but not there yet and it took 10 months to get us here.

The second generation (bitfury/knc) are running at 1.0-1.6W per GH/s. That would move the needle if EVERYONE converted over to 3.75 to 10x more efficient. That would only bring it to 16PH/s for network total in however it takes for people to convert to more efficient units.

There are devices out there that are promising 0.6W per GH/s for release in 1Q 2014, and will again be subject to delivery issues, but if everyone moves to the new efficiency, we're talking another 2.6x jump at the worst, so 41.6PH/s.

In other words, there is no way that the network will reach 120PH in April, let alone 2014, without some big surprise event like - cost of power going down by 1/3 globally (not happening), a viral growth of participation in bitcoin (barrier to entry is still pretty high right now) or the announcement of a 0.2W per GH/s chip (which is also unlikely because the 0.6W per GH chips are a 28nm process and Moore's law would dictate that we're not tripling our efficiency in 5 months)