5 ms·

It will almost certainly cost more money to obtain 51% hashrate than you could gain from a short position, by the time mining rewards aren't enough of an incent

It will almost certainly cost more money to obtain 51% hashrate than you could gain from a short position, by the time mining rewards aren't enough of an incentive. Even the NSA doesn't have enough computing power to achieve 51%.

tlrobinson 13y ago
I agree with the first sentence, but I'm pretty sure the NSA could do it. Until recently it would have only cost a few million dollars to fab enough ASICs for a 51% attack. I haven't seen a recent analysis, but it's probably still only in the 10s of millions (possibly 100s of millions)
- ewillbefull 13y ago
  The NSA probably specializes in general purpose computational power, they couldn't afford to compete with ASICs at this point. But they probably could destroy any alt-coin alternative to bitcoin in a heartbeat, especially Litecoin which is designed to be mined by general purpose hardware.
  - dragonwriter 13y ago
    > The NSA probably specializes in general purpose computational power, I would agree that this is likely in the sense that the bulk of their expenditures on computing power are likely directed at general purpose computing power. That doesn't really mean that they couldn't have expended fairly vast amounts on special purpose computing, as well. > they couldn't afford to compete with ASICs at this point. Given that the NSA is kind of concerned about monitoring things (especially things used by people who are particularly concerned about avoiding government monitoring), and given the degree to which people with that interest have been the focus of bitcoin, on what basis do you assume that they didn't get in on the ground floor with ASICS as they became available (or earlier)?
    - tlrobinson 13y ago
      ASICs (or mining in general) aren't needed to monitor all Bitcoin transactions. A mining node is indistinguishable from a non-mining (full) node. The only reason for ASICs is mining (either legitimate or as an attack)
  - johnnyanony 13y ago
    >The NSA probably specializes in general purpose computational power, they couldn't afford to compete with ASICs this is an absurd statement because the us military, nsa, and cia design their own custom chips and have a black budget to finance the operations and some of the most talented and experienced mathematicians, scientists, and programmers to back them up. when the central banks actually find bitcoin to be a threat they will crack the encryption, do a 51% attack, or conduct some other discrediting attack on its validity and security. I am short term bullish on bitcoin though I am hoping for the best for btc or other decentralized digital currencies.
kevinpet 13y ago
This is naive. There's no natural process that ensures the network stays well distributed. If Bitcoin takes off, then why would the next hardware miner maker sell the product? Maybe they find it better to just run a significant portion of the market rather than deal with clients. 10 years from now, what if we have a handful of bitcoin mining firms, similar to the current role played by market makers like Getco or Knight? You can't just dismiss this point. You need to come up with some process that makes it more valuable for someone to be a small player in the network than to go all-in.
dragonwriter 13y ago
> It will almost certainly cost more money to obtain 51% hashrate than you could gain from a short position If true, that means it wouldn't be an action of a rational attacker bent on financial gain, but not all attackers are rational, and financial gain isn't the only conceivable motive for an attack on bitcoin. > Even the NSA doesn't have enough computing power to achieve 51%. How, exactly, are you assessing the quantity of the relevant type of computing power available to the NSA in order to make this claim?
- ewillbefull 13y ago
  NSA builds general purpose computing in massive scales. The bitcoin network is beyond the scope of all of the world's supercomputers combined. I don't have some irrational belief that the NSA has a bunch of ASIC SHA256 devices laying around. They're too late.

tlrobinson 13y ago

I agree with the first sentence, but I'm pretty sure the NSA could do it. Until recently it would have only cost a few million dollars to fab enough ASICs for a 51% attack. I haven't seen a recent analysis, but it's probably still only in the 10s of millions (possibly 100s of millions)

ewillbefull 13y ago
The NSA probably specializes in general purpose computational power, they couldn't afford to compete with ASICs at this point. But they probably could destroy any alt-coin alternative to bitcoin in a heartbeat, especially Litecoin which is designed to be mined by general purpose hardware.
- dragonwriter 13y ago
  > The NSA probably specializes in general purpose computational power, I would agree that this is likely in the sense that the bulk of their expenditures on computing power are likely directed at general purpose computing power. That doesn't really mean that they couldn't have expended fairly vast amounts on special purpose computing, as well. > they couldn't afford to compete with ASICs at this point. Given that the NSA is kind of concerned about monitoring things (especially things used by people who are particularly concerned about avoiding government monitoring), and given the degree to which people with that interest have been the focus of bitcoin, on what basis do you assume that they didn't get in on the ground floor with ASICS as they became available (or earlier)?
  - tlrobinson 13y ago
    ASICs (or mining in general) aren't needed to monitor all Bitcoin transactions. A mining node is indistinguishable from a non-mining (full) node. The only reason for ASICs is mining (either legitimate or as an attack)
- johnnyanony 13y ago
  >The NSA probably specializes in general purpose computational power, they couldn't afford to compete with ASICs this is an absurd statement because the us military, nsa, and cia design their own custom chips and have a black budget to finance the operations and some of the most talented and experienced mathematicians, scientists, and programmers to back them up. when the central banks actually find bitcoin to be a threat they will crack the encryption, do a 51% attack, or conduct some other discrediting attack on its validity and security. I am short term bullish on bitcoin though I am hoping for the best for btc or other decentralized digital currencies.

kevinpet 13y ago

This is naive. There's no natural process that ensures the network stays well distributed. If Bitcoin takes off, then why would the next hardware miner maker sell the product? Maybe they find it better to just run a significant portion of the market rather than deal with clients. 10 years from now, what if we have a handful of bitcoin mining firms, similar to the current role played by market makers like Getco or Knight? You can't just dismiss this point. You need to come up with some process that makes it more valuable for someone to be a small player in the network than to go all-in.

dragonwriter 13y ago

> It will almost certainly cost more money to obtain 51% hashrate than you could gain from a short position If true, that means it wouldn't be an action of a rational attacker bent on financial gain, but not all attackers are rational, and financial gain isn't the only conceivable motive for an attack on bitcoin. > Even the NSA doesn't have enough computing power to achieve 51%. How, exactly, are you assessing the quantity of the relevant type of computing power available to the NSA in order to make this claim?

ewillbefull 13y ago
NSA builds general purpose computing in massive scales. The bitcoin network is beyond the scope of all of the world's supercomputers combined. I don't have some irrational belief that the NSA has a bunch of ASIC SHA256 devices laying around. They're too late.