Because they are literally selling you a machine that just sits around and 'makes' money.
If the amount of profit from using these machines was greater than the profit from selling the machines.. then the company selling the machines wouldn't sell them to you..
>If the amount of profit from using these machines was greater than the profit from selling the machines.. then the company selling the machines wouldn't sell them to you
Unless they couldn't make the machine without you paying for it first. Or they don't want to break the system by reaching >51% of the network (after which the value of btc plummets down immediately, thus making the machines worthless).
If a single group controls more than 51% of the mining power they can cheat the system (double spending, things like that). If such a monopoly was created people might lose their trust on Bitcoin and try to leave the system.
A lot of people will tell you, "Think of the upfront costs" and that sort of thing. My opinion however is that it's nothing we haven't seen before. Maybe the customer makes money in the long run. Maybe it takes 6 months for the unit to pay for itself. The manufacturer on the other hand gets their profit up front. They don't have to wait 6 months, and they don't have to worry about whether the value of bitcoins will crash.
Certain profit today, vs. less certain profit of greater magnitude tomorrow. Seems familiar.
There is only a certain amount of BTC to be had every hour. If they brought these in by the bucketload they could dominate the mining utterly. They would then also effectively control the network people may lose trust, BTC value could tank etc etc.
In a recent interview (the one that photo is drawn from), that CEO guy acknowledges that from a purely selfish standpoint, shipping the ASICs at all has a high opportunity cost, but http://motherboard.vice.com/blog/engineering-the-bitcoin-gol... it was more honest to ship:
> This is an obvious question, but why aren’t you mining your own chips?
> Like I said, the current reality is vastly different from what we originally had planned. We always figured we would be late to the party. Butterfly Labs and bASIC (another ASIC developer that ultimately failed to launch) were supposed to launch in succession. But due to various complications and delays, it didn’t work out that way.
>
. If they had delivered on their promises, they would have shipped months before our first batch. We wanted to start selling chips so people could make their own units, providing a hedge from a single entity becoming too powerful, and then move onto a new project. That was our main goal. We wanted to prevent this potential monopoly. As it turned out, we became the monopoly we tried to prevent.
> You aren’t doing any mining at all?
> Nope. Fun fact: none of the Avalon team have their own mining units (outside of test units)... We’ve had plenty of opportunities, such as keeping the technology to ourselves and simply mining, but that was never our intention. So how much is that trust worth? We think it’s worth $5000 by itself. Beyond that, the calculation is based on the current difficulty rate so that break-even point occurs in a month. Which, by the way, as an investment vehicle is a fairly unique proposition. It should be mentioned also that the cost of the batch two was also 75 bitcoins. That the exchange rate has moved is out of our control. And in the end, this was never what we wanted. If we wanted to maximize our profit, we could charge much more. For us, this has always been an ideal-oriented problem, not a business-oriented one. This is one of the primary reasons we aren’t clearly in the black. We’re lucky that we held onto the second batch of bitcoins because if it weren’t for the appreciation, we would be very, very deep in the red, which ultimately still isn’t that big of a deal.
(And presumably, they're airmailing the components. Why risk a boat?)
At the current exchange-rate, it seems counterproductive for Avalon to sell their units instead of running them in-house to recoup the cost of production (in a couple of days), and then continue to generate profits.
Also, since they are the only game in town (plus BFL & private players probably), they could take advantage of their position in the short term, before the difficulty increases by another order of magnitude.
But Avalon has pre-sold them, presumibly to fund the development & construction. To keep the machines, they'd need investment money, which they might not have.
I'm not sure if they're actually selling new ones, or just shipping preorders from 6 months ago. And the reason they took preorders was to get the capital to fund the production in the first place.
It does seem like it would be awfully tempting to delay those preorder deliveries a little bit and use the machines first.
Yes. Luckily, Avalon has a philosophical committment to improving the bitcoin transaction network.
Other suppliers which haven't established this idea can be accused of receiving orders and just running them locally for a while before sending them out.
Currently the supply is limited to 3600 per day, so if they did take all their orders and 'premine' for a day, they could at a maximum make about a half a million dollars a day (at the current trading rate). Note that if they started selling half a million dollars a day worth of bitcoin, they'd see the prices drop to less than $10, making their daily profits to around 50k. At that point, it may be more profitable just to ship their product. If you did buy one of these ASIC miners, most people don't realize that as soon as customers get ahold of them, the difficulty is going to go way up, since the hard limit of 3600 coins would be reached earlier in the day if it didn't. So, if they had 1000 customers, the most each customer could get in a day would be 3.6 coins... which at today's rate is $468 per day. Not too bad, but what would it cost to get that kind of hashrate? If it's $10k, then you'd have to mine for 21 days nonstop just to have it pay for itself. That's assuming the difficulty level would stay the same, which it probably wouldn't since the other players (BFL, etc) would release their products, thus decreasing your daily haul. Also, if everyone who mined coins sold, you'd look at a depreciating value for the currency, so your 3.6 coins might only be worth half as much.
Are you sure? I was pretty sure that miners and wallets were distinct, and wallet programs are the ones that participate in the distributed transaction engine.
Mining verifies all transactions in the network. That's the whole point of the system, making sure that transactions can't be faked. In order to fake a transaction, you would need to control 51% of the total hash rate of the Bitcoin network, which is impossible.
Sure, but I'm pretty sure that the only reason you need to keep adding and adding ever more capacity is to prevent someone else coming along and doing the 51% attack, right?
Other than that, adding capability doesn't benefit the network, it burns a lot of power as people effectively just burn electricity to try to increase their share of a fixed payout.
> I'm pretty sure that the only reason you need to keep adding and adding ever more capacity is to prevent someone else coming along and doing the 51% attack, right?Other than that, adding capability doesn't benefit the network, it burns a lot of power as people effectively just burn electricity to try to increase their share of a fixed payout.
Sure, if you want to efficiently allocate resources, all bitcoin miners could get together as a cartel and limit the overall hash rate to 1Mhash/hr, and collectively the bitcoin transactions network would consume a lot less power.
However, the temptation would always be there to add a little extra hash power to the network and network to claim more of the mined blocks and transaction fees. And pretty soon you'll be back where we are today, with each person acting in their own selfish interest, addiing as much hashing power as they can afford to run to get as big a share of a fixed pie as they possibly can.
> In order to fake a transaction, you would need to control 51% of the total hash rate of the Bitcoin network, which is impossible.
Relevant to the article that this thread is linked to, it wouldn't be at all impossible if you are the single high-rate producer of by far the most efficient mechanism for mining bitcoins, and you really wanted to disrupt bitcoin.
(It might be economically irrational for anything that seems to a sane person to be a plausible assignment of utilities, but then, lots of socially destructive acts that actually occur in the real world meet that description.)
"In order to fake a transaction, you would need to control 51% of the total hash rate of the Bitcoin network, which is impossible."
Impossible?! The claims about why this polynomial-time attack on Bitcoin is not really a problem seem to get more outrageous with each passing day.
I find it interesting that you simply assume that there is no other attack on Bitcoin. There is no proof of this. There is not even a good enough definition of the security goal of Bitcoin to allow for such a proof.
Call it economically improbable. Right now you could maybe get 51% with a 10-20 million dollar investement in ASICs. If you took control of the network and started forging transactions, bitcoin would no longer be useful and all those coins you stole wouldn't be worth very much.
EDIT: I vaguely remember the Linode hacker mention that a more reasonable attack possibility would be if you slipped an backdoor into the popular bitcoin client libraries, and then it could be used to steal.
Maybe you only want to deny confirmations to your enemies. I can imagine that the US government might want to prevent people from making donations to Wikileaks. Maybe someone just wants to destroy Bitcoin, for reasons that are beyond your imagination.
Really though, trying to figure out your attacker's intentions or goals, or their willingness to commit resources to an attack, is a fool's errand. This is even more so when your system has known and exploitable weaknesses. The Germans learned this lesson the hard way with Enigma:
I guess that by "Godwin's Law of Nazi Analogies" this discussion has officially run its course :P.
I think I agree with the major implications of what you are saying: bitcoin hasn't demonstrably withstood nation-state levels of attack effort, so calling anything impossible is premature.
If the bitcoin network survives for long enough [which can't be taken for granted] this will eventually happen, but it hasn't demonstrably happened yet.
'this' is in many ways a harbinger of the future. Maybe (probably, though it has lasted longer than many predictions, including mine) it will be bitcoin probably it won't be but the future is definitely going to include something like bitcoin.
Imagine the same about newspapers: This is really perverse, some of our best writing talent got pulled into writing news articles in order to get people to read the ads.
Advertising is a means, not an end and as such it is a pity that so much talent gets devoted to getting people to click on it but that will burn itself out somehow and then we're left with the byproducts.
> Imagine the same about newspapers: This is really perverse, some of our best writing talent got pulled into writing news articles in order to get people to read the ads.
This isn't an apt analogy. Ads may bring in the revenue but people are doing something useful. Same could be said for Google---ads bring in the revenue but the product (i.e., search) is still highly useful.
The computational power used to mine bitcoins are doing absolutely nothing of useful value to society. I believe there was an HN discussion earlier about tying this in with Folding@Home or SETI@Home and make the mining process a bit more productive.
Except that miners aren't actually breaking any laws. But more importantly will be to see what how quickly the market floods with new coins and the impact that has on collectors.
I'm not sure I follow, Bitcoins are mined at a rate of one block every 10 mins (on average). The difficulty setting adjusts based on how much processing power is thrown at the problem. Supply is constant.
If the difficultly increases after the coins are mined, then it would always be playing catch-up and actually a flood could occur.
It is my understanding (but I could be mistaken, there is a lot of FUD around BTC) that someone could in theory mine 500 hashes in a ten minute window then 'retire' having boosted the mining rate. This would not only net them a lot of bitcoins immediately but also restrict others supply of bitcoins.
Supply is roughly constant given a slowly increasing mining pool, supply isn't shock resistant.
I imagine they use a decent amount, but not nearly as much as would be required to run a cluster of computers to mine a comparable quantity. I think with an ASIC, electricity costs are negligible compared to the output.
Keep in mind that your return is based on the current net hash rate of all mining computers, everywhere. Combined with the 6+ month wait time, you can buy a $20,000 rig that would mine you $500/day today but by the time you get it the hash rate will have risen so high that you'll be lucky to out-mine the cost of your power consumption.
I wouldn't be surprised if these companies were using each rig for a couple months in their prime before shipping them out.
I am surprised the article does not mention that Avalon, in addition to ready-to-use devices, just started selling bare mining chips to OEM manufacturers last week. And they already sold $5M worth of them (500 thousand chips) in the past 8 days alone! This equates to 140 Thash/s (twice more power than the current network). Lead time is 9-10 weeks. Manufactured on-demand at TSMC.
Unless I've interpreted BTC wrong - this means that as soon as they go live, the current miners are going to see their returns drop by 2/3? And if many are already at the edge of profitability in terms of electricity/BTC, they'll drop off the network?
The BTC network may be utterly dominated by ASIC players within a few weeks.
Are they any more energy efficient than racks of GPUs?
A decent GPU would still be plenty profitable even after a 3x difficulty increase. For example the HD 7970 (about $400) currently mines $130/month, with electrical costs around $15/month assuming a worldwide average rate of $0.10/kWh, making a $115/month profit. After a 3x diff increase, this would turn to $45/mo revenues and $30/mo profits.
However this 140 Thash/s of Avalon chips is just a fraction of what is about to hit the network. All ASIC vendors combined, plus future Avalon deliveries will push the difficulty far more than 3x higher in the next 4-6 months.
Current ASICs do 150-200 Mhash/Joule (at 12V). Much better than GPUs topping out at 3-4 Mhash/Joule (at 12V).
I would think they are sophisticated enough that they could have a parameter for their client specifying a target hash rate and slowly ramp it up (over weeks/months), looking more or less just like other sorts of hardware coming online. It would take some work to hide in a pool or whatever, but I don't think it would be a big problem.
You can't hide it to the point where you'll be able to mine at a rate high enough to make back the investment on the hardware development unless you do it openly. It's an economic argument, not a technical one.
56 comments
[ 2.8 ms ] story [ 103 ms ] threadIf the amount of profit from using these machines was greater than the profit from selling the machines.. then the company selling the machines wouldn't sell them to you..
Unless they couldn't make the machine without you paying for it first. Or they don't want to break the system by reaching >51% of the network (after which the value of btc plummets down immediately, thus making the machines worthless).
They take pre-orders for one batch of machines. They build that batch of machines. They soak-test those machines, and then start sending them out.
Now, with the money from the pre-orders and from the soak-test bitcoins, they can build a second batch which they keep.
Certain profit today, vs. less certain profit of greater magnitude tomorrow. Seems familiar.
You could make a fortune at the factory just doing the burn-in test for 24 hours.
In fact I am curious why they would even ship them from China, you'd lose money the weeks on the boat, just plug them in at the factory...
> This is an obvious question, but why aren’t you mining your own chips?
> Like I said, the current reality is vastly different from what we originally had planned. We always figured we would be late to the party. Butterfly Labs and bASIC (another ASIC developer that ultimately failed to launch) were supposed to launch in succession. But due to various complications and delays, it didn’t work out that way. > . If they had delivered on their promises, they would have shipped months before our first batch. We wanted to start selling chips so people could make their own units, providing a hedge from a single entity becoming too powerful, and then move onto a new project. That was our main goal. We wanted to prevent this potential monopoly. As it turned out, we became the monopoly we tried to prevent.
> You aren’t doing any mining at all?
> Nope. Fun fact: none of the Avalon team have their own mining units (outside of test units)... We’ve had plenty of opportunities, such as keeping the technology to ourselves and simply mining, but that was never our intention. So how much is that trust worth? We think it’s worth $5000 by itself. Beyond that, the calculation is based on the current difficulty rate so that break-even point occurs in a month. Which, by the way, as an investment vehicle is a fairly unique proposition. It should be mentioned also that the cost of the batch two was also 75 bitcoins. That the exchange rate has moved is out of our control. And in the end, this was never what we wanted. If we wanted to maximize our profit, we could charge much more. For us, this has always been an ideal-oriented problem, not a business-oriented one. This is one of the primary reasons we aren’t clearly in the black. We’re lucky that we held onto the second batch of bitcoins because if it weren’t for the appreciation, we would be very, very deep in the red, which ultimately still isn’t that big of a deal.
(And presumably, they're airmailing the components. Why risk a boat?)
Also, since they are the only game in town (plus BFL & private players probably), they could take advantage of their position in the short term, before the difficulty increases by another order of magnitude.
It does seem like it would be awfully tempting to delay those preorder deliveries a little bit and use the machines first.
Other suppliers which haven't established this idea can be accused of receiving orders and just running them locally for a while before sending them out.
Other than that, adding capability doesn't benefit the network, it burns a lot of power as people effectively just burn electricity to try to increase their share of a fixed payout.
Sure, if you want to efficiently allocate resources, all bitcoin miners could get together as a cartel and limit the overall hash rate to 1Mhash/hr, and collectively the bitcoin transactions network would consume a lot less power.
However, the temptation would always be there to add a little extra hash power to the network and network to claim more of the mined blocks and transaction fees. And pretty soon you'll be back where we are today, with each person acting in their own selfish interest, addiing as much hashing power as they can afford to run to get as big a share of a fixed pie as they possibly can.
Relevant to the article that this thread is linked to, it wouldn't be at all impossible if you are the single high-rate producer of by far the most efficient mechanism for mining bitcoins, and you really wanted to disrupt bitcoin.
(It might be economically irrational for anything that seems to a sane person to be a plausible assignment of utilities, but then, lots of socially destructive acts that actually occur in the real world meet that description.)
Impossible?! The claims about why this polynomial-time attack on Bitcoin is not really a problem seem to get more outrageous with each passing day.
I find it interesting that you simply assume that there is no other attack on Bitcoin. There is no proof of this. There is not even a good enough definition of the security goal of Bitcoin to allow for such a proof.
EDIT: I vaguely remember the Linode hacker mention that a more reasonable attack possibility would be if you slipped an backdoor into the popular bitcoin client libraries, and then it could be used to steal.
edit2: spelling.
Really though, trying to figure out your attacker's intentions or goals, or their willingness to commit resources to an attack, is a fool's errand. This is even more so when your system has known and exploitable weaknesses. The Germans learned this lesson the hard way with Enigma:
https://en.wikipedia.org/wiki/TICOM
I think I agree with the major implications of what you are saying: bitcoin hasn't demonstrably withstood nation-state levels of attack effort, so calling anything impossible is premature.
If the bitcoin network survives for long enough [which can't be taken for granted] this will eventually happen, but it hasn't demonstrably happened yet.
Imagine the same about newspapers: This is really perverse, some of our best writing talent got pulled into writing news articles in order to get people to read the ads.
Advertising is a means, not an end and as such it is a pity that so much talent gets devoted to getting people to click on it but that will burn itself out somehow and then we're left with the byproducts.
This isn't an apt analogy. Ads may bring in the revenue but people are doing something useful. Same could be said for Google---ads bring in the revenue but the product (i.e., search) is still highly useful.
The computational power used to mine bitcoins are doing absolutely nothing of useful value to society. I believe there was an HN discussion earlier about tying this in with Folding@Home or SETI@Home and make the mining process a bit more productive.
It is my understanding (but I could be mistaken, there is a lot of FUD around BTC) that someone could in theory mine 500 hashes in a ten minute window then 'retire' having boosted the mining rate. This would not only net them a lot of bitcoins immediately but also restrict others supply of bitcoins.
Supply is roughly constant given a slowly increasing mining pool, supply isn't shock resistant.
http://techland.time.com/2011/05/23/report-police-confuse-bi...
I wouldn't be surprised if these companies were using each rig for a couple months in their prime before shipping them out.
Looking at the global hash rate it's clear that this is not happening. (Before anyone asks, you cannot hide mining.) http://bitcoin.sipa.be/
https://bitcointalk.org/index.php?topic=181982.0
http://store.avalon-asics.com/?product=avalon-asic-chips-100...
The BTC network may be utterly dominated by ASIC players within a few weeks.
Are they any more energy efficient than racks of GPUs?
Vastly. GPU's are space heaters compared to ASICs and FPGA's.
However this 140 Thash/s of Avalon chips is just a fraction of what is about to hit the network. All ASIC vendors combined, plus future Avalon deliveries will push the difficulty far more than 3x higher in the next 4-6 months.
Current ASICs do 150-200 Mhash/Joule (at 12V). Much better than GPUs topping out at 3-4 Mhash/Joule (at 12V).
The key questions to predict ASIC mining profitability is:
- how much will be invested in ASIC hardware
- future prices of bitcoin
- how long will it take for the ASIC hardware to roll out
The only "easy-money" I see here is if you have access to ASIC mining hardware _right now_.
I would think they are sophisticated enough that they could have a parameter for their client specifying a target hash rate and slowly ramp it up (over weeks/months), looking more or less just like other sorts of hardware coming online. It would take some work to hide in a pool or whatever, but I don't think it would be a big problem.