No no, the real bloodbath will start when there will appear ASIC not supported by community... Then investors will make bloodbath to such ASIC producer - because hashing ASIC would be worthless... So there will be collective motion towards hashing algorithm change... Remember, users first, miners next... BitCoin was not for miners... There were multiple clones "for miners", and where are they now ? ;-))))

No matter how you envision it, its the reality of ASIC development: (very) high NRE with nearly negligible per chip, and therefore per GH, costs. Since the bitcoin mining market is mostly limited by how much bitcoins are worth, you get a very weird situation.

lets put it this way; a processed wafer costs on the order of a 1-2000 dollar and should provide on the order of a terrahash or more worth of asic hashing power; actual numbers dont really matter here. Ignoring the other costs (PCBs etc ) which are relatively minor,  a single batch of 12 $1000 wafers could potentially provide as much hashing power as the entire network, and therefore be worth millions today.

But if you sell the equivalent of hundreds of wafers, its not going to net you hundreds of millions, the market just isnt there for that, it will still net you on the order a few million dollar.  So if you can generate a comparable revenue by selling several dozens of devices or several thousands of them, which makes most sense?  You go with the former obviously, anyone would. Not that it would prevent you from selling thousands later on.

So anyone buying that early terrahash of asic power would get what will probably seem like a great deal; yet he could lose most of his investment 6 months later just because you or a competitor decided to crank up the production volume almost arbitrarily. Suddenly his 1TH is no longer 5 or 10% of the network but 0.05% and it will only take a relatively minor investment by you or a competitor to produce those extra chips.

So this is not criticism of BFL, its just the logical consequence of a bitcoin ASIC that can generate virtually unlimited hashrates while the total profit miners can generate is fixed. As a result, a miner his ROI is completely dependent on your, or a future competitors production volume. Knowing the variable cost of production is virtually nothing, thats pretty damn risky. So I seriously think some sort of assurance that you will not produce more than x TH over y time would be a very good idea to reassure early customers.

I think what ngzhang is saying is "seeing is believing". One cannot dispute the fact that a 1.05G/20W/$499 device was (pre)announced, and in the end turned out to be a not-quite-as-svelte 0.832G/80W/$599 device.

I have been disappointed by the Singles' real specs relative to the inflated hype and anticipation, but have purchased them in spite of that and in spite of the delays and have been more than satisfied.

So based on that history, for the upcoming (pre)announcement of the SC, I'll remain cautiously optimistic but take any announced specs and shipping dates with the prerequisite grain of salt. Looking forward to more detail.

There's no contradiction. It is are users, who are willing to use BitCoin who by-fact pay compensation to miners for the network security. Not vice-versa.

So open-asic would work, then everybody will know exact takeover cost, while NDA-ASIC won't work. So BFL would likely end up in condition, where they would have to sell these chips very close to silicon cost, and opening specs OR OTHERWISE - they would become private owners of overall BitCoin system security. As it was correctly said, that if you count cost of silicon wafers - that are trivial costs to outperform current network hashing power.

Then - when someone privatizes BitCoin - it may be good or not... For example - if I would privatize it, and control majority, what I would do ? Ask yourself that question. Because actually I will become LIABLE for all damages that users could get by my actions, but of course there's no legal framework about it. And if I start with mining only purpose - then - I will be screwed - because nobody wants "just mining", people need goods and services paid in secure way. So model would change from trust-nobody decentralized to trust-one who has majority, not much difference compared to webmoney then. And whether it will work or not - it depends on actions of that de-facto owner.

And then it comes to BFL.......... and trust..... would I trust my assets to them with their approach ? Ask this question to yourself... And ask another question - who you know here on forum to whom you would trust say $10k holding in the system, and knowing that he say can dump $100M cap to $50M for own profit on exchange rates and won't do that ?

What I definitely know that greedy people would screw this thing likely, as they will not pass any opportunity to make money, even at delicate moments, where making money puts risks on all users... Even worse if they will silently start ripping off users... Then bankers would just laugh at it and say - damn internet crooks... So that 20W 1Gh/s and other "pre-announce" buzz definitely shows how you can trust your "rulers of BitCoins" :-) Banking payment systems is all about trust and then people behavior...

DUN DUN DUNNNNNNN

I am a happy BFL customer.

2 questions:
1)  Where can I sign an NDA to get more information?
2)  Can I order yet?

Thanks, just looking for a ballpark figure.

Music to my ears

Well done BFL, glad to hear you're still innovating.   Sorry to see the same sad old trolls up to their own tricks (you'd think, after they were humiliated when you came through the first time around, they'd have learnt their lesson, but there's no accounting for stupidity I suppose).

As for the competitors getting scared and starting attacking BFL, I guess you're scared for a reason hey.   Here's an idea, let your product stand on its own merits without the underhanded attacks.