FPGA, but not ASIC. Litecoin runs a version of Scrypt with the parameters tweaked so that an ASIC could easily blow all existing miners out of the water. If they would have kept the scrypt using the large amount of memory, it would have been better.

You could form an intermediary on GLBSE to gather money to buy shares indirectly.

This makes me not feel so good about my recent BFL purchase.  

I hate to say it, but I do agree that a centralization like this could severely damage Bitcoin in a scenario I imagine:
- If enough hashing power is generated by vlad, and the price/BTC stays the same, then most GPU miners will drop out.
- If vlad sells all of his generated BTC to cover costs and investor dividends, then the price will drop further, as more BTC is sold instead of traded for goods or held.
- If the price drops, and ASIC mining increases, then difficulty is going up while price per coin is going down.  More miner dropouts.
- More price drops, as vlad ends up with a higher percentage of mined coins, continues selling them all, and confidence in BTC wanes.
- Eventually, vlad owns 70-80% of hashing power, because the price has dropped so far that only those with ASICs or free electricity are profitable to mine.
- Extreme loss of confidence from everyone, since one person controls the Bitcoin network.  No one wants to risk having money held in BTC, so price drops to $0.25/BTC or less.

I doubt Bitcoins will ever cease to be used, but a death-spiral like this could definitely happen if ASICs aren't released to the general public.  One person/company having ASICs while no one else does would be disastrous.  The barrier to entry is too high for another person/company to risk spending the time and money on developing, but no one else will be able to compete without ASICs.  If no one else can compete, no one else will mine.  If no one else mines, no one else gets coins to spend.  The bitcoin economy would be slowed, and confidence in the project is destroyed.

Vlad, I very much encourage you to make ASICs publicly available.  It is in your own best interest that you do so.

TBH this would not surprise me.

It is either ArtForz or the Spanish folks in Grenada or him.

Not good for BTC in any case ...

i thought an order of magnitude was 10x generally, so that would be 20-30 times more efficient than a gpu wouldn't it?

vlad, what protections do investors have that their funds won't be used to subsidize the R&D, only to have the principles of the company accumulate personal mining hardware on the cheap, after investors have paid for the high entrance barrier to ASIC development?

Monopolies and oligopolies lead to collusion. If you sell shares in a public corporation you owe your shareholders a return on investment, and these returns are judged by the marketplace on a quarterly basis under the theory of unlimited growth. Of course in the real world, no matter how large it may appear, there are limits to growth. So over time those monopolies or oligopolies particularly those addicted to oil (which after all powers the mining rigs) will pathologically seek out new ways to increase (make those) quarterly profits. Sure that might lead to green energy, but of course it is just as likely to create new ways to extract monopoly rents from bitcoin users and the bitcoin supply chain. At present, unfettered capitalism with out broad based representation (open source/distributed) has spawned monster corporations acting as sociopaths. This is not seen as theory, rather the question is what to do about it. Harnessing the wealth of many for the benefit of a few without any input or oversight by those who are being harnessed.  What is Facebook.

Unless bitcoin users have some sort of participatory powers in decisioning how bitcoins are administered, we will be replicating the current monetary system albeit without inflation.

Order means power.

So 2 orders of magnitude is 10^2 = 100x.  3 orders of magnitude = 10^3 = 1000x

One of the things I always try to be sure I understand before I invest is -- Who is the customer?  In this case it isn't very clear...

It is easy to get tied up in the exciting technological aspects of ASIC and their impressive efficiency but the fact remains that for an enterprise to be an ongoing concern they are either subsidized or have customers that pay for products or services...

you prompted me to search online, and the accepted definition seems to say an order of magnitude is 10 times the original, 2 orders of magnitude 100 times the original and 3 orders of magnitude 1000 times (as you said).

i guess it was the first order of magnitude i had stuck in my head, although each oom is simply another increase by 10x, i just wasn't compounding the increases by factor of 10.

so 100 to 1000 times more efficient is the actual range quoted by the OP.  seems an awfully large range of performance estimate if they are so far along in the design.

If Vladimir (or anyone invested in Bitcoin) reaches 51% he is unlikely to make the 51% 100% by rejecting all other blocks because that would invalidate the entire currency and the customers (users) will switch to some other crypto currency.

I for one do not fear a monopoly.

Btw monopoly is a natural by product of using a free-market structure. Nothing lasts forever though and so too monopolies will die.

"This too shall pass" - King form Persian Fable