7 ms·

At the very end, Kaminsky says that large financial actors and nation states have the ability to deploy massive mining farms, and that somehow this is a bad thi

by mrb 14y ago

At the very end, Kaminsky says that large financial actors and nation states have the ability to deploy massive mining farms, and that somehow this is a bad thing. I do not think so. These actors would merely compete between each other, thereby enhancing the security of the Bitcoin network, making it harder for an individual actor to perform majority attacks ("51% attack") and rewrite the block chain. A successful attack, with large actors already participating in the mining industry, would require collusion, which I think is unlikely. Perhaps Kaminsky simply meant that deploying massive mining farms would help them profit from Bitcoin more efficiently than the average user. This would be false. Contrary to other industries, there is no economy of scale in Bitcoin mining. I witnessed this first hand when expanding my GPU ops from 2 Ghash/s to ~60 Ghash/s in 2011. Indeed, the smaller you are, the more overhead costs become negligible and eventually effectively "free". The student mining in his university dorm, or the individual mining on his mini desktop ASIC or computer at work, have zero mining costs when operating: free electricity, free hosting, free A/C, free network connectivity. However, the ones managing huge farms have to pay for data centers, maintenance technicians, electricity, etc. This is why Bitcoin mining is a long-tail system: most of the mining power comes from a large number of small-time miners. (I should add that at this moment, in 2013, there is a small window of opportunity for a large actor to dominate mining. ASICs are barely starting to hit the network, so in theory a large actor could deploy many of them and represent more than half of the mining capacity. However this window is rapidly closing. Most people estimate the network hash rate is going to grow by 10-50x in the next 12 months.)

peterjancelis 14y ago
The risk is not in governments profiting from adding hashing power to the network, the risk is in them obtaining a majority of the hashing power and doing a 51% attack.
- feral 14y ago
  As an additional point to this, something I often don't see acknowledged is that ASICs have to be produced by chip fabs, which due to their nature are capital intensive and centralized, and hence relatively easy to bring under government control. I guess that as a result, it'll always be possible for governments to get majority hashing power, by taking control of where the ASICs are produced.
  - peterjancelis 14y ago
    If you are afraid of ASICs you should look into Litecoin.
  - zerohp 14y ago
    The NSA already has their own chip fab.
tlrobinson 14y ago
I recently calculated the cost of 50% hashing power in ASICs, and it was something like $1M. Even if the hash rate grew 10-50x it would cost less than $50M. Well within reach of governments or large corporations. That said, as adoption and thus the value of Bitcoin increases, the incentive to mine and thus secure the network rises, so there's a nice feedback loop.
- NewAccnt 14y ago
  Rumor has it that Butterfly has been sitting on their preorders and using it to build a massive ASIC fleet. Whether or not these are true, as you suggest, building a 51% fleet is not out of reason if you take the right path and keep your mouth shut. There has already been forks of the blockchain due to powerful miners. http://mineforeman.com/2013/03/14/what-the-fork-was-that-a-forking-post-mortem/ http://mineforeman.com/2013/03/14/what-the-fork-was-that-a-f...
  - mrb 14y ago
    These forks are not caused by powerful miners. They are caused by miners mining 2 blocks almost at the same time. And the fork of March 14 was caused by a bug.
- mrb 14y ago
  Firstly, your numbers are off. If the hash rate grew by 50x, we would be at 3.5 Phash/s, so the attacker would need to deploy another 3.5 Phash/s to attack the network. And Butterfly Labs's ASIC price is $50 per Ghash/s. So it would cost $175M. Secondly, you cannot buy $175M of ASICs from BFL; they are too small of a company. Thirdly, by the time you design, build, and deploy ASICs yourself (12 months+[1]), the network would have grown again, maybe by another 2x/5x/10x, who knows... you would have needed to account for this by spending respectively $350M/$875M/$1.75B ! So effectively by the time the network reach 3.5 Phash/s, it will be too late. [1] For comparison it took more than a year for the DOE to deploy the #1 supercomputer, Titan, out of commodity hardware.
  - Retric 14y ago
marshray 14y ago
> These actors would merely compete between each other https://en.wikipedia.org/wiki/Bretton_Woods_system https://en.wikipedia.org/wiki/Bretton_Woods_system
bcoates 14y ago
They are not obligated to play by the rules and fairly compete with each other; if their goal is not to acquire bitcoins by mining but to disrupt the system by publishing problematic blocks, they don't raise the difficulty factor. There are absolutely economies of scale for mining (it's just making and running custom hardware, a classic economy of scale business), it's just the combination of low profitability and a small market [1] means it's not worth doing unless you're getting paid to attack the network. [1] at the very peak pre-crash price, $341 million/year of revenue for mining every single block.
- mrb 14y ago
  Imagine a graph where the x-axis represents the size of your mining farm (from small to big), and the y-axis the percentage of the mining income spent on operating costs. For small values of X the curve is close to zero. As we move to the left the curve suddenly jumps, and then decreases and decreases but without ever coming back close to zero. This illustrates my point: small miners have operating costs close to zero, above a certain size you suddenly start needing to spend quite a bit on operating costs (first rack, first A/C unit, etc), and as you scale up and up yes there is some amount of economy of scale, but never enough to reduce your operating costs close to zero. I say "economies of scale do not exist" but what I mean more precisely is that they are not sufficient to cause big miners to push small miners out of the market.

by mrb 14y ago

peterjancelis 14y ago

The risk is not in governments profiting from adding hashing power to the network, the risk is in them obtaining a majority of the hashing power and doing a 51% attack.

feral 14y ago
As an additional point to this, something I often don't see acknowledged is that ASICs have to be produced by chip fabs, which due to their nature are capital intensive and centralized, and hence relatively easy to bring under government control. I guess that as a result, it'll always be possible for governments to get majority hashing power, by taking control of where the ASICs are produced.
- peterjancelis 14y ago
  If you are afraid of ASICs you should look into Litecoin.
- zerohp 14y ago
  The NSA already has their own chip fab.

tlrobinson 14y ago

I recently calculated the cost of 50% hashing power in ASICs, and it was something like $1M. Even if the hash rate grew 10-50x it would cost less than $50M. Well within reach of governments or large corporations. That said, as adoption and thus the value of Bitcoin increases, the incentive to mine and thus secure the network rises, so there's a nice feedback loop.

NewAccnt 14y ago
Rumor has it that Butterfly has been sitting on their preorders and using it to build a massive ASIC fleet. Whether or not these are true, as you suggest, building a 51% fleet is not out of reason if you take the right path and keep your mouth shut. There has already been forks of the blockchain due to powerful miners. http://mineforeman.com/2013/03/14/what-the-fork-was-that-a-forking-post-mortem/ http://mineforeman.com/2013/03/14/what-the-fork-was-that-a-f...
- mrb 14y ago
  These forks are not caused by powerful miners. They are caused by miners mining 2 blocks almost at the same time. And the fork of March 14 was caused by a bug.
mrb 14y ago
Firstly, your numbers are off. If the hash rate grew by 50x, we would be at 3.5 Phash/s, so the attacker would need to deploy another 3.5 Phash/s to attack the network. And Butterfly Labs's ASIC price is $50 per Ghash/s. So it would cost $175M. Secondly, you cannot buy $175M of ASICs from BFL; they are too small of a company. Thirdly, by the time you design, build, and deploy ASICs yourself (12 months+[1]), the network would have grown again, maybe by another 2x/5x/10x, who knows... you would have needed to account for this by spending respectively $350M/$875M/$1.75B ! So effectively by the time the network reach 3.5 Phash/s, it will be too late. [1] For comparison it took more than a year for the DOE to deploy the #1 supercomputer, Titan, out of commodity hardware.
- Retric 14y ago

marshray 14y ago

> These actors would merely compete between each other https://en.wikipedia.org/wiki/Bretton_Woods_system https://en.wikipedia.org/wiki/Bretton_Woods_system

bcoates 14y ago

They are not obligated to play by the rules and fairly compete with each other; if their goal is not to acquire bitcoins by mining but to disrupt the system by publishing problematic blocks, they don't raise the difficulty factor. There are absolutely economies of scale for mining (it's just making and running custom hardware, a classic economy of scale business), it's just the combination of low profitability and a small market [1] means it's not worth doing unless you're getting paid to attack the network. [1] at the very peak pre-crash price, $341 million/year of revenue for mining every single block.

mrb 14y ago
Imagine a graph where the x-axis represents the size of your mining farm (from small to big), and the y-axis the percentage of the mining income spent on operating costs. For small values of X the curve is close to zero. As we move to the left the curve suddenly jumps, and then decreases and decreases but without ever coming back close to zero. This illustrates my point: small miners have operating costs close to zero, above a certain size you suddenly start needing to spend quite a bit on operating costs (first rack, first A/C unit, etc), and as you scale up and up yes there is some amount of economy of scale, but never enough to reduce your operating costs close to zero. I say "economies of scale do not exist" but what I mean more precisely is that they are not sufficient to cause big miners to push small miners out of the market.