The more people get excited about faster ways to hoard bitcoins and eventually make a nice profit (presumably in USD in the end) the dramatically more cynical I become about its future. When the bitcoin discussion was mainly around its use as a currency for things like silkroad, VPNs etc I actually thought it seemed to have a promising future. I really hope I'm wrong, but I have a feeling in 2023 the bitcoin discussion will be relegated to nostalgic jokes about how crazy we used to be.
This is working out exactly as planned - the security of the system against double-spending is predicated on honest nodes controlling the majority of the hashing power in the bitcoin network, so the fact that people are working very hard to add hashing power to honest nodes is desirable for the system as a whole.
Indeed, it is pretty crazy. I know someone with an Avalon ASIC who claims Satoshi is a genius. I think he's biased by the fact he has a small device that plugs into the wall and makes him more money than his day job.
> The more people get excited about faster ways to hoard bitcoins
People seem to be oblivious to the fact that a currency must be available in order to be useful. There is a reason that Gresham's Law has achieved "Law" status.
Oddly, the right thing to do for currency acceptance would be to shutdown all of the ASICs.
I think people are more oblivious to the fact that a "deflationary currency" is inherently flawed because it by definition results in hoarding, which is not what you want people to do with your currency. Why would you bother loaning or investing it if you can make more money just sitting on it?
If USD was limited to currency in circulation then it would not be nearly as inflationary as it is.
Once an established lending system based on bitcoins is established it will have the same problems as our current currencies. It wasn't printing money that caused the real estate bubble, it was the way to banking system has the power to expand credit and the systemic risk that expansion of credit causes.
Once there are fractional reserve bitcoin banks it will no longer be a deflationary currency.
Will people want to borrow in defalationary biased currency though? Better to borrow in Bernanke Bucks where there is an unstoppable torrent being emitted monthly.
IMHO the Real estate bubble was from the Fed bailing the prior tech bubble by suppressing interest rates - which encourages borrowing. When loans are paid back, the currency created by taking the loan is extinguished but they have been juicing the system by continually suppressing rates via QE/bond purchases and not letting the real market decide the risk (ie: price of bonds). Now that interest rates have hit the floor - the next bubble to pop is the USD itself.
If the USD pops then BTC may be less deflationary. The bubble did not pop because mortgages were paid back.
Loans are generally repaid in currency, paying back a loan creates even more currency as the additional currency created by the interest becomes an asset against which loans can be issued.
Deflation is not desireable in itself, but it is not a crippling disadvantage. Eg Japan has had a very long deflation spiral and it didn't affect the usability of their currency. And the west has been living in a world of negative real interest rates since the 2008 crisis.
As long as we don't have a big bitcoin credit market and we don't want to regulate the bitcoin "printing press" to adjust for market cycles, it won't even be a disadvantage. = As long as it's only a secondary currency to most users.
I'm not a gold bug and I see many problems with using gold as a currency, namely that the amount of currency in circulation should depend on how much stuff we can dig out of the ground. (In the case of BTC something about how many times a SHA-256 hash can be calculated.)
The idea that gold fell out of use as currency because of industrialization is either a completely new discovery in economics, or completely wrong. I'm pretty sure it's the latter.
I don't think the Bretton woods system counts. That's between countries, I'm talking about everyday preference for gold by individuals. I suppose I should have said specie instead.
Yeah but that I define as "monetary metal for Central Bank settlements" gold, not "day-to-day normal-people transactional currency". Was it still "money" / monetary after 33? Certainly, until '71 (42 years ago, not 39). But "currency" (circulating to facilitate/lubricate individual micro rather than macro exchange&trade)? Hardly ;) Also not "legal tender" -- as in, for courts enforcing contract liabilities or tax payments.
TL;DR: only outside the US, only to Central Banks.
I doubted that these ASIC rigs would ship any time soon. Wouldn't it be more profitable for the vendors to simply to delay/cancel orders and run the rigs 24/7 themselves?
The profitability of ASICs is going to decrease at a rapid rate once they start shipping to consumers. Additionally, in a few months the market is going to be flooded with cheap ASICs.
From BFL's perspective if they don't release first they're dead in the water. They don't know where their competitors are compared to them but they recognize it's a winner take all situation. The natural thing to do would be to take pre-orders to fund the manufacturing and start realizing profits now. Also, if your competitors are offering pre-orders you should be too.
The evolution was CPU -> GPU -> FPGA -> ASIC. Now that we're at the ASIC stage growth is going to slow considerably. There's no longer order of magnitude gains to be made.
>at the ASIC stage growth is going to slow considerably
Yes, buy these are 110nm v1 designes, it looks like they will be able to squeeze something like 10x per watt by moving to better process and iterating on design. And probably additional 10x in cost by ordering more than current quantities that are in the ballpark of the smallest possible order.
Right now entire network runs at speed equal to all hardware in first two Avalon batches (300 and 600 units). When batch 2 and 3 and BFLs hit in the future, everything else is going to be wiped out. But then it will be wiped out again by the next iteration.
Difficulty adjusts week-to-week. Being the only guy with an ASIC is like a license to print money. Being the 10,000th person with an ASIC will be like trying to mine with a netbook today.
Supply will always be constant however how it's divided changes -- people who previously enjoyed lack of competition are now having their feet held close to the fire.
I see it as a futures contract on the value of the BTC. Selling a mining ASIC locks in profit for the seller, betting that this will exceed the time-value generated from the rig itself. Buying a rig locks in a loss, betting that it will pay for itself. We can't be certain whether the BTC will be at US$1,000 or US$1 a year from now. You could make a case for either side, that's why the contract exists.
Which, by the way, doesn't mean they are betting against BTC. They just elected to take sure profit up front, over less certain profit of potentially greater value long-term.
> I doubted that these ASIC rigs would ship any time soon. Wouldn't it be more profitable for the vendors to simply to delay/cancel orders and run the rigs 24/7 themselves?
Only if they're "long BTC". What makes you think they are? They're just providing shovels & tools to "miners in a gold rush". Historically, such providers have always done phenominally well.
At least they get the rigs at production cost to run them---selling them for good ol' USD may just be a hedge or sideshow, only time can tell.
There is energy expended during the process though. I imagine if this continues 'banks' of the future will be data centers. Also might be a way to make money on idle servers at data centers.
If anything this is an extremely interesting time in currency experimentation. Side effect of this is massive energy cost, another awesome side effect is hardware/supercomputer miniaturization and evolution. Great for really hard problems like verification (as in mining here) and maybe brain research/processing or other really hard data crunching problems.
I'm amazed that there hasn't been a scandal with some Googler or Microsofty or Appler or academic finding a way to use some fraction of their unused cycles for mining.
Their competitor Avalon sells the units for "the first 30 days profit of the device". So whatever the current price of Bitcoins and mining difficulty affects the price.
I'm sure the same thing was said during the gold rush days and it was the store owners that sold prospecting equipment that made some very stable money.
what makes you think they aren't doing just that? they could start shipping them out and make a profit on the machines just as the difficulty rate picks up and there is more profit to be made in selling miners rather than mining.
Part of the value of Bitcoin (if any) is that it's decentralized. It should be more profitable to sell the miners out than to hoard them because if they're distributed the cryptographic safety of Bitcoin is greater.
Decentralization boosts miners' profits by making bitcoin sounder, and therefore making its price higher.
Miner's profits have to do with the stability of the network and the demand for bitcoins. At high bitcoin prices, their return is better than when the prices are low, other things (thing, mining difficulty) being equal. It's much like gold mining: it has become much more profitable of late because the difference between extraction price ($250 to $400 per ounce in most places) and sale price (roughly $1,500 per ounce) is how much the miner takes home.
Decentralization increases this price because it makes bitcoin sound. If there's no one door to knock down, or one person to subpoena, the cyberpunk currency can robustly survive losses in its network and still have enough computing power trained on it to prevent attacks.
From what I understand these devices started shipping right after the BitCoin peaked... Maybe the vendors kept "testing" them for just a bit longer than they expected?
How can one litigate against an entity beyond the jurisdiction of the US court system, e.g. say the rig manufacturer was the crook and not in the United States? What are the options there?
The options are to sue them in their foreign country or sue them in the United States. If you sue them in the US, you face jurisdictional issues (do they have sufficient contacts in the US to make a lawsuit here fair?) as well as judgment recognition issues (you would probably have to attach the damage award to a contract they have with someone in the US to collect if the foreign country court system doesn't recognize the US judgment).
Liquidation: Corporate governance issues are a matter of the country where the defendant is organized. So if the law there allows them to do it, they can. I'm less familiar with the process of chasing the money. Trying to remember back to law school here: In the US, the debt doesn't just disappear. It would either attach to the old entity (which would have to be paid a fair price for the acquisition of its assets) or the new entity (which acquired its assets and liabilities per the acquisition agreement). So if it has no assets to begin with, it's possible. If it has real assets, it's more difficult to do. Not impossible, just more difficult.
Also 1/5th the energy efficiency that BFL originally promised when taking orders, and their most vocal spokesperson claimed that there was no way that they'd be significantly more power-hungry than predicted, that anyone who suggested otherwise was trolling, and that the only reason to draw attention to the fiasco when their FPGA mining hardware which was announced with impressive efficiency and blew out the power budget so badly they had to redesign the boards was to shill or troll. Right up until they blew their power budget and had to redesign the ASIC boards.
Additionally, he tried to convince people to buy BFL rather than one of the competing boards by arguing it was so much more power efficient miners using them would be forced to shut down. It's barely more power efficient than ASICs which shipped months ago, and those companies are already looking at newer models.
Oh, for what it's worth that's also less than 10x the power efficiency of the best FPGA boards...
I estimate it'd produce around 0.25 BTC/day, mining in a pool. That's about $30/day USD at today's price.
As ASICs come online, the difficulty is going to go up, which will reduce the daily generation rate for any given miner. I've been trying to decide whether or not to order a device now based on my profitability prediction for months from now when I might actually get it. I'm guessing by the time I get one I might be able to produce at least 0.05 BTC/day. Assuming that's $6/day, the $274 device will take about six weeks to pay for itself. Not too bad, but I don't know if keeping tabs on it afterwards is worth it for $6/day. It'd only be worth it if I think BTC are going to be worth a lot more in the future... getting back to hoarding coins as an investment rather than using them as currency.
They probably missed the clock-rate by a factor of 2x, or decided to use half the number of chips and overclock them to save money. It may also have been their backup plan because people are willing to pay 8x as much for power, but they're less willing to get 1/2 the hashrate.
If you run a CPU at the same voltage, but at half the speed, it uses 1/2 the power because CMOS circuits only really use power as the transistors change state. However you're also able to run at about 1/2 the voltage without getting errors. Because V=IR, 1/2 the voltage at the same resistance is 1/2 the amperage, and since P=IV, that's 1/4 the power for normal circuits. Of course CMOS circuits already run at 1/2 the power. In the real world there's also leakage and such, but in a perfect world doubling the clockrate takes 8 times the power.
Transmeta's Crusoe was the first processor to really take advantage of this, by varying the voltage and running slower to save battery life. Before this time (2000), laptops would use a duty cycle to save power. Soon all the major chip manufacturers started copying them.
I ordered the 60gh/s SC Single the night preorderes opened like a year ago. I think there are maybe ~7 orders before me. I hope I get it soon. I paid 220 btc@5.45 usd for it, let's see if its worth it.
I ordered a Jalepeno almost immediately. I chose to pay by wire, and the email i received included "Your order position won't be penalized for any time it takes to send you this payment information. Please send me an email with a copy of the wire confirmation form. Once payment has been received, I'll respond with final confirmation of your order."
I still haven't paid and am going to now. I wonder if they will honor it or not.
"time it takes to send you this payment information. "
I don't see where it says you won't be penalized for the time it takes for you to make a payment. With that said - I'm curious, did people pay in advance of shipping?
If he paid with Bitcoin, it was probably handled through BitPay, which might have in turn converted it to USD before paying it out to the company. I'm not sure the company could have foreseen the spike in BTC and held off the USD for that long.
I'm sure with enough digging around I could find this out for myself, but could someone familiar with things comment on how the performance of these boxes compares to a standard ATI card miner?
High-end ATI cards will do < 1GH/s, with most midrange cards below 500MH/s. This does 5GH/s, while using an essentially an order of magnitude less power than a PC-based rig will use.
The more important specification is the MHash / Joule (also found in the link below). The video says this thing does 5 GHash/s @ 30W. That's 100x the efficiency of the best videocards (1000x if you count the power consumption of the rest of the PC).