6 ms·
Doing a 51% attack on a major coin, and actually profiting from it, is much more complicated than this chart implies. Here are some of the problems you'll run i
by Meekro 8y ago
Doing a 51% attack on a major coin, and actually profiting from it, is much more complicated than this chart implies. Here are some of the problems you'll run into:
1. NiceHash doesn't have anything like the hashpower you need to attack a major coin such as BTC or ETH. The chart admits that NiceHash has only 2% of the capacity you need to accomplish this on Bitcoin. You'll need to start buying ASIC hardware or graphics cards to cover the other 98%. Both of those are hard to obtain in very large quantities quickly.
2. Since you have to buy the hardware, you can expect to pay much more than the NiceHash prices imply. Bitmain is one of the major suppliers of such hardware, so let's use their prices as an example. One of their top Bitcoin ASICs is a 14 TH/s unit that costs about $1,000 [1]. So you need about 2.5 million of these units at a cost of 2.5 billion dollars. Not that Bitmain has the capacity to fill such an order.
3. The cost of those ASICs is just the beginning. Have you seen photos of those Chinese data centers that have racks of mining ASICs on shelves? Each of those data centers has maybe a few percent of Bitcoin's hash rate, so you'll need to build something at least 30x larger. Your electric bill alone will exceed the GDP of some small countries. Since you've already spent billions of dollars on ASICs, hopefully you have some money left to pay it and hire a data center ops team.
4. How fast can you build all this out? Bitcoin's hash rate has grown 30% over the last month[2]. You might sink billions of dollars into this project only to discover that you've come up short.
5. Congratulations, you can now make a few million dollars by 51% attacking the blockchain. You would, for example, deposit some bitcoins with a bunch of exchanges, sell them, withdraw dollars, and then use your hashpower to unwind those deposits and put the bitcoins back into your own wallet. It'll take a few hours (or less) for every exchange to notice what you're doing and freeze withdrawals. The bitcoin price will free-fall as everyone wonders about how to prevent this next time. You may or may not be arrested and charged with wire fraud. You're done!
[1] https://shop.bitmain.com/product/detail?pid=00020180523201006625r2KQCnC10650 https://shop.bitmain.com/product/detail?pid=0002018052320100...
[2] https://blockchain.info/charts/hash-rate https://blockchain.info/charts/hash-rate
- machinesmachine 8y agoYeah, this isn't the last season of Silicon Valley, you've explained it really well. Theoretically it's very dangerous vector of attack, however, in practice seems impossible seeing how big Bitcoin is over all.
- JumpCrisscross 8y agoYou wouldn’t attack Bitcoin by buying hardware. You’d steal it. An industrial-scale bot et, for example.
- Meekro 8y agoA massive bitcoin mining virus wouldn't help. Let's say there are 2 billion PCs in the world, as this random site estimates[1]. Let's say each of them has a high-end graphics card.. say, a GTX 1060 that can do 20 MH/s. Let's say your virus can break into every single computer in the world, and run all those graphics cards at 100% without being noticed. Now your mining power is: 20 MH/s * 2 billion = 40000 TH/s, or 0.1% of Bitcoin's hash power. [1] https://www.quora.com/How-many-PC-exist-in-the-world https://www.quora.com/How-many-PC-exist-in-the-world
- JumpCrisscross 8y ago> 2 billion PCs Hence my specification for industrial scale. More likely: a corrupt data centre system admin or government official. In any case, if you’re taking on the Bitcoin network, doing it with fairly-obtained resources would be silly.
- arisAlexis 8y agodata center can't compete with asic mining farms
- stale2002 8y agoI think the implication would be that you steal the asic mining farms. IE, you somehow compromise 51% of the EXISTING hashpower, do your attack. It doesn't matter to YOU that their mining hardware will become worthless.
- wafflebear 8y agoIt seems very implausible. Anyone with that level of sophistication may as well look for exploits in bitcoin exchanges and steal the bitcoins directly.
- auganov 8y agoAnd if you try to attack one of the smaller coins, odds are it'll just get forked. Which is a risk with big coins too, but at least there's a chance the original chain retains some value.
- mrep 8y agoNot sure what you mean? Wouldn't the double spender immediately tumble them by exchanging them for other coins across a bunch of different exchanges making a fork to fix it nigh impossible?
- auganov 8y agoMost smallcaps have tiny volumes and aren't traded on many exchanges. Getting the coins into (and out of) an exchange isn't immediate. Even worse most of the volume is likely to be on a trustworthy exchange that does some basic KYC (and extended KYC for withdrawals). Basically, for most $1M cap coins, you'd be lucky to get $5k out if all goes well. I mean, I guess it's still viable. It's just street-robbery-viable.
- btown 8y agoThe interesting thing is that the Chinese government could theoretically nationalize all those data centers and be very close to (or far beyond) 51% on any proof-of-work-based cryptocurrency [0]. This would, of course, be a one-way street, immediately lead to a price crash, and likely be unprofitable... but it might be viable, say, as part of a scorched-earth cyberattack to cause chaos and wipe out international wealth parked in Bitcoin. It's heartening to see things like Ethereum moving in a proof-of-stake direction [1], and there's a lot of hope for a lot of the new cryptocurrencies out there. But Bitcoin itself is far more centralized and government-controlled than a lot of people think. [0] https://medium.com/@homakov/how-to-destroy-bitcoin-with-51-pocked-guide-for-governments-83d9bdf2ef6b https://medium.com/@homakov/how-to-destroy-bitcoin-with-51-p... [1] https://www.ccn.com/ethereum-moves-one-step-closer-to-proof-of-stake-as-casper-receives-first-release/ https://www.ccn.com/ethereum-moves-one-step-closer-to-proof-...
- wpietri 8y agoGiven that money laundering is a significant use of Bitcoin, and given China's extensive history with capital controls [1] and their willingness to execute people for corruption [2], it would not shock me much to see them do this for purely internal reasons. Chinese leaders are historically not big fans of things they can't control. [1] It's a hot topic: https://www.google.com/search?q=china+capital+controls https://www.google.com/search?q=china+capital+controls [2] E.g.: http://time.com/4298731/this-is-how-much-money-you-can-take-in-bribes-before-the-chinese-authorities-execute-you/ http://time.com/4298731/this-is-how-much-money-you-can-take-...
- Jommi 8y agoWhy would people use Bitcoin for money laundering?
- wpietri 8y agoI don't understand the question. Bitcoin's aim was to be international electronic cash. People use actual cash for money laundering. Why wouldn't they use electronic cash? If you're interested in learning more, I'd suggest starting with Google: https://www.google.com/search?q=bitcoin+money+laundering https://www.google.com/search?q=bitcoin+money+laundering
- SilasX 8y agoYeah, for profit miners have a built in incentive against doing such an attack, since it destroys the value of their own assets. But there’s still a danger from a well funded government or corporation willing to take a loss to destroy one of the currencies.