796 at least has a "money service permission" license. Whether thats enough, I dont know  (actually doubt it very much), but that is completely irrelevant. The issue is the assets themselves. Fabrizzio and theswede solicited investments in an unregulated security from  EU and US investors. HK laws almost certainly dont allow this either, but even that doesnt matter/ Its not because you solicit investments in a nigerian scam company that its legal here just because the supposed company is in Nigera. If you cant go after Fabrizzio, you certainly can go after theswede. Thats assuming anyone knows who he really is.

You dont really believe that, do you? Even the IPO was illegal. I know nothing of Chinese law, but it was certainly illegal in the US and EU and almost certainly in China too. You cant just publicly solicit money (or btc) for tradable shares in a company; thats an unregistered security. Thats why BTCT is closing.

moreover, even if the IPO somehow were legal, management cant just close down a company and claim all the IPO funds were spent without proving that. There is no way all coins could have been spent. Lets begin by seeing those "100K chips". Lets see some bookkeeping and invoices. If they cant provide that, its fraud, and even in China thats not legal.

Care to give some more insights in to the cost structure?

For the asic, allow me to make a guestimate based on cointerra's specs.
3 100mm² dies in a package yields 502-720 GH /chip
Lets go with the low estimate gives 166GH per die.

A 300mm wafer yields ~600 10x10mm² candidates or 100TH (at 100% yield).  AFAIK, processed 28nm wafer cost less than  $4000 in volume. Low volume runs may add significantly to that, if you have to go through to an intermediary.
Let be generous and for easy math put it $10,000 per wafer if you include yield, slicing, packaging, handling,.. thats ~$100/TH.

Now thats only silicon cost, excluding NRE and of course there are other costs, like PCB, assembly, housing, cooling, PSU, etc, but I find it hard to understand how over time, you would not be able to sell at (much)  less than $500/TH. Am I way off in my silicon cost estimate?  Or what else is so expensive? PCBs?

Only 2 of them "samuel" and none of them with qualifications or even geographical location that could possibly match.

Probably descendants from the well known royal Noi family in Nigeria.

Seriously though, if that is your guy, you are screwed.
Alternatively, If the only guy with some knowledge of whats going on, the "nominee director"  of an IT company without website is not on linkedin and no one knows who he really is, you are probably screwed too.

I will tell you what the name reminds me off.  Wild conspiracy with zero evidence to back it up, but here you have it:  Chaang Noi. aka Goat.

Its the only samuel noi on linkedin aside from this one:

http://gh.linkedin.com/pub/samuel-mensah-noi/77/b66/333

Is this samuel noi ?

http://gh.linkedin.com/pub/samuel-noi/55/11/b72

Almost, but not quite. I see no reason to believe BFL is mining significantly with customer hardware. If they were, difficulty would be way higher.
There is a much better reason to delay shipping; as long as you dont ship, difficulty stays low and the perceived value of your products remains higher, particularly if you dont tell anyone how much TH you sold already.
(Its the same ploy as preoders, get people's cash months in advance without telling them how much you sold and thus how much difficulty will increase)

Had BFL shipped all their orders in a reasonable timeframe, they would have gotten far far less (pre)orders or at much lower prices. Thats their prime motivation for going slow.

No but you could say the speed increase we've enjoyed over the past decades will hit a brick wall. In so far it hasnt already. And moore's law is running out of steam too. IIRC on some chips we are down to feature sizes that are just a dozen atoms wide, transistor size is approaching atomic levels. It just doesnt get smaller than that.  We are not there yet, still a few shrinks that are probably feasible but the end is in sight.

Buying coins has nothing to do with it (btw, you would buy them at todays price, but not at $1? I sure as hell would).
And you may shut down your inefficient miners, but even at $1/BTC today those 28nm asics would still earn more than they cost in electricity.
Now $1 is pretty extreme, but you get my point.

Nor can you predict it. Whats going on now has nothing to do with BTC price and BTC price will have next to no influence on network hashrate over the next months either; even if it goes to $1 or $10000, those asics will be deployed and cant be deployed faster than they are anyway. BTC price will only determine at what level the growth will taper off.

according to posted IRC chats, TheSeven wasnt even aware there was a 130nm chip, has never spoken to anyone working on it and knows nothing about it. He is also more a software guy than anything else.
But yeah, he'll fix it overnight, especially when the only guy who really knows the asic is gone.

Meanwhile theswede says "they will keep adding hash rate". and we are supposed not to laugh? The only way hashrate is going to be added is when their bitfury's finally show up.

A few years ago, we have seen far steeper increases in difficulty than what we see now. You cant spot that  in the current curve anymore as it has become noise, but here is a good chart for you:


http://thegenesisblock.com/bitcoin-network-reaches-1-petahash-per-second/

These werent "bubbles" and nor is the current rise,  its the network adjusting to new technological realities.  This apparent exponential growth wont last, just like the previous ones didnt last, but that doesnt mean it will implode, just that it will taper off once the transition is complete. FWIW, whats happening now has been predicted many years ago, it was inevitable we would move to asic's at some point given they are inherently vastly superior in GH/$ and GH/W than FPGA, GPU's and CPU's. If anything this transition is happening much slower than I would have predicted.

But we are now in the last phase. After asics (and particularly 28nm asics)  there is nothing else to create this kind of network growth. Or if there is, its yet to be invented. The network will continue to grow to 10+ PH without slowing down, probably even accelerate further in the coming months. Then it will begin to slow, but it wont stabilize before we reach somewhere between 50 and 500PH (assuming constant btc price).
Economics dictate that.

I guess some people here want Ken post in this thread every day even if its just feeding trolls and posting non news ("wafers at fab moved from machine A to B", "exciting new step today I cant announce due to NDA") or make bold promises just to boost up the share price for the daytraders.

Others may prefer he spends his time more productive and make sure this company can actually deliver.

$500+ per GH doesnt exactly sound like a steal either does it?

BFL isnt the only one shipping, Bitfury's are selling like hot cakes too.
As for BFL doing this on purpose.. perhaps. Artificially keeping difficulty low would have given them a lot longer to sell more preorders at higher prices. It would also cost them customers and credibility but Im quite willing to assume overall its a net win for them. Then again there is Hanlon's razor: Never attribute to malice that which is adequately explained by stupidity. Or in this case, incompetence. Im just on the fence whether or not incompetence alone explaines this kind of shipping clusterfuck.

Interesting point. That would be a huge worry for me if I were selling mining hardware. But arguably there is a provision for this in the legislation.


Thats from the EU directive:
http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:1997L0007:20071225:EN:PDF

If that were to come to court, I think it could swing either way.

ah shoo, why did you have to tell them?
Now they will stop perordering all those "ROI" generating miners that will secure our network.


Well, that is if they understand " =/= "

Horten Ho 229
(dont ask me why its posted there)

How well does this model work when you retroactively apply it to the periods where we transitioned from CPU to GPU or FPGAs ?

I appreciate you put a lot of work in to this, but building mathematical models to make predictions that dont factor in facts like that $10M's worth of next generation hardware is due to be delivered over the next few months, makes it usability rather questionable.  Its not like charts can predict if KnC will ship all that its promised in 2 weeks or wont ship anything at all, and that alone would roughly double difficulty.