Bitcoin Forum

Is this chart correct? It would show how improbable something like that would occur to Bitcoin, even to a government. And yet, if the attack succeeds, it can end up without profiting.

But I found it curious that some other currencies seem to have a relatively acceptable possibility. And like the shitcoin BGold, they are open to such an attack. This shows how risky it is to bet on bad projects or badly made copies. We're not in a casino, be careful where you invest.

https://pbs.twimg.com/media/DeQVA15U0AE-mTE.jpg

Source: http://rindex.io/attack-cost

Roughly so. If you compare the calculated hashrate against one of the more popular ASICs on the market, the figure stands at about 2 billion. However, this assumes that suddenly half of the network right now goes rogue and decides to attack it. If not, you would need more than 2 billion dollars since the hashrate needed has to be 51% of the total hashrate (after the attack starts).

Its rather resource consuming and it isn't that big of a problem since none of the government really cares that much to spend their budget on attacking a cryptocurrency. You probably wouldn't profit all that much when you attack; doubt exchanges credits more than 2 billion dollars worth of Bitcoin automatically without raising a red flag.

That's assuming the attacker decide to do 51% attack by purchase lots of ASIC/GPU, but it's not really accurate number because :
1. There's no way you could get thousand of ASIC/GPU units easily
2. Even if they could, i'm sure the price of ASIC/GPU will be more expensive due to high demand and low supply
3. Unless you have connection, i doubt you could use lots of eletricity and big mining location easily

I think adding backdoor to ASIC, take control over mining poll or release Botnet malware are cheaper/easier solution for the attacker.

This is generally not correct as the attackers will not be acknowledging most or all of the blocks of honest miners, so the difficulty will not increase. It is in the same vein as the selfish mining problem.

I'm not really sure but it was just a thought. But take for example, the network has 1000GH/S of hashrate and you are planning to execute a 51% attack so you're adding 510GH/S of hashrate. However, the total hashrate is now 1510GH/S and 510/1510 of hashrate isn't 51% of the total network.

If you don't have 51% of the whole network (ie. more hashrate than the others), you don't have a 100% chance of executing an attack. The honest chain still has a chance of overtaking you and you have to acknowledge their blocks. The probability of success decreases with the number of blocks.

You're correct that the probability of success is less than 100% without a true 51% as you described, and my description was not accurate either. But in summary:

http://hackingdistributed.com/2013/11/04/bitcoin-is-broken/

"But the outcome is that the selfish mining pool, on the whole, nullifies the work performed by the honest pool through their revelations. [...] But this threshold is non-existent in the current implementation -- selfish mining is immediately profitable. Our proposed fix raises the threshold to 25% if universally adopted. And, while there may be other fixes, no fix can raise it above 33%. So, at least 2/3rds of the Bitcoin miners have to be honest."

A lot of disruption can be caused with a 33% attack. Attacks with repeatable attack surfaces (such as depositing, withdrawing, then depositing again to an exchange) will eventually succeed. But they definitely won't be able to double spend at will or prevent the network from operating normally.

Exactly. The over $2 billion number is just today's economic cost of owning 51% of the hash rate. For any new entrant to want to take over, the "cheapest" would probably be simply to introduce new hash power... about 101% of all current existing hash power. This would then take this simple estimation to about $4.5 billion, today. even for a 33% attack as Ix suggests, you'd need the full 51% of all current (to own 33% of all)... so that is a very expensive "disruption".

I'm saying introduction of new machines is only possible as it would be far too costly and difficult to attempt to take over all existing power... and you couldn't fail taking over a single machine.

There is no need to buy ASICs, in the current state were almost all mining power is concentrated within less than ten pools that's enoght to 'buy' a pool admin(s).

but also a "pool" is not a "farm" it doesn't have any ASICs it is just a server that the miners with their ASICs connect to. if a pool is corrupted the ASICs will connect to another one and hashrate of that that pool drops.

But would it be possible to "rent" some pools for a short period of time, I imagine it would take some time, more than 1 hour, to redirect these ASICs., and use a dormant farm for an attack?
How big would your farm need to be if you could rent BTC.com and AntiPool, which at that time represent 39%, for only 40 minutes?

Why would even mining pools agree to "rent" their hashpower? If they actually did that, they would quickly lose their users. I don't really think if there would be many farms interested in this form of attack. Long-term earning is much wiser than trying to cheat and sell as much as possible. The rest of the network would quickly realize that the 51% attack was performed. Just look what is going on with Verge which is/was being attacked. The 51% attack is the reason why there are people who think that banning ASICs and mining only with GPUs is a good way to prevent it from happening. And... Bitcoin Gold was also attacked.

What about the recent 51% attacks on Bitcoin gold etc?  Does that make an attack on Bitcoin more feasible?

No, it doesn't. Bitcoin Gold is a fork of Bitcoin which is completely different and independent. Bitcoin Gold can be mined with GPUs whereas Bitcoin needs to be mined with ASICs so there is no competition between these two coins (in terms of mining). Bitcoin Gold was supposed to be more decentralized thanks to the different mining algorithm but it turned out to be a huge flaw. Bitcoin developers have no reason to get back to GPU mining, it would be disastrous since the number of ASICs is enormous.

Why will they connect to another pool if 'the corrupted' one gives them more reward (keep in mind they wil not only gain from double spending but also bear the all blocks' reward).

They won't gain from double spending. The block rewards will still be the same, unless the pool owner decides to incentivise them with the profits gained from the double spends. In fact, they could be losing out. Depends on how the community reacts to it, the dishonest chain could be abandoned by them and any block rewards would be gone. Participating in attacks against Bitcoin is essentially equal to them giving the support to kill it. Keep in mind that they own ASICs and they can only be used to mine SHA256D coins. If anything, they would be against these forms of attacks.

I would say that they probably won't have time to switch to another pool by the time the double spends happens.

That is quite interesting, does anyone know the cost for Verge as they just dealt with the issues? Thanks

The attacks on verge were not real 51% attacks as far as I know. The attacked exploited a glitch in the code around block timestamps which they were supposed to have fixed. In basic english the guy was messing around with the times that blocks were created to earn free verge.

Nope.
You miss the point.
'The corrupted' pool must override N blocks mined by some-other-pools with his own blocks -> hence the reward for those N blocks shall also be grabbed by 'the corrupted' pool.
In other word, this kind of attach just naturally combines 'double spending' and 'selfish mining'.

Just wondering , how come robustness index of ether come greater than BTC when there hash rate is less than BTC?

For somebody coming outside of the network need to bring  102% of hashrate.
If network has 100 machine/miners, then outsider need to bring 102 machine/miners so that more than 51%  (102/202) share in network can be gained to execute the attack.

not worth it at the moment,no 51% attack is going to be financially profitable at the current state of things
but the less adequate mining pools there left,or the better some of them become compared to other
it is very possible that 3-4 of the major pool admins could conspire and pull something like that off

https://i.imgur.com/wSlao9Nl.jpg

today,three top pools can do the 51% attack on their own, 54.3% of the total network hashing power  at their disposal

AntBleed  (https://www.antbleed.com/)is already a thing, so it wouldn't surprise me if a government would be able to bribe Bitmain. Assuming large miners did not mitigate AntBleed, a 51% attack would be possible today given that Bitmain can shut off a large amount of miners, and they control a large amount of Bitcoin's hash rate in their pools. BTC.com is around 20% of the network, and AntPool is another 16%.

Hey, I found another article relating to this, in this one it assumed that the attacker just buys the hashpower from Nice hashable, which would reduce overhead costs massively. With Bitcoin Gold and Verge having some trouble with attackers lately I would love to know if these figures are accurate.

https://www.crypto51.app/

Looks accurate for me. NiceHash has 84.09 MSol/s (https://www.nicehash.com/algorithm/equihash) that could be rented for Equihash (though it would cost quite a bit of money), and BTG has a total network hashrate right now of just over 23 Msol/s (https://bitinfocharts.com/bitcoin%20gold/), so someone could rent from Nicehash and become 75% of the BTG network.

Verge uses Scrypt, and has 2.38 TH/s on the Verge network (https://whattomine.com/coins/150-xvg-scrypt). NiceHash has 20 TH/s that is rentable (https://www.nicehash.com/algorithm/scrypt), so with enough money (renting 20 TH/s for 1 day costs ~15BTC), someone could become 90% of the Verge network.

Thanks for the insight, seems very interesting to me. Do you think this might have an effect on POW coins with low hashrate to switch to something like proof of stake or other alternatives to ensure security in their network?

The scenario of someone outright buying new miners for hashpower is unlikely. Who would sell and buy that many miners. That can happen if their is some,; to use that popular world, collusion between bitmain and Governments. Such a collusion is always a possibility but I don't think this can be done without someone noticing. And if the word gets out, then a PoW change is the easiest way to render all those ASICs redundant.

The other two scenarios are:

1. Pools coming together to attack
2. Bitmain flipping the kill switch.

Now the first one is unlikely because such a step will only hurt their investments by denting bitcoin's reputation. Bitmain flipping the kill switch and outright attacking the network as a last resort to prop up BCash is the most likely scenario.
This is scary as Roger, CSW and Bitmain have continued their tirade and marketing against bitcoin in mission mode. It is strange that there isn't more discussion on this in the forum. Maybe it is time to take this to the Ivory Tower.

Hey, I found another article relating to this, in this one it assumed that the attacker just buys the hashpower from Nice hashable reducing costs massively. It wouldn't work for POW coins with massive hashpower such as BTC but it would work for smaller coins.

https://www.crypto51.app/

Ethereum uses a completely different algorithm as compared to Bitcoin. Ethereum's algorithm was pretty much more resistant (not anymore lol) to ASICs as compared to Bitcoin. Everyone can mine Ethereum with profit. Obviously, it has changed now but the technology appears to not be as mature and the cost of the ASICs is pretty high.
That is, of course if they want the attack to last for an extended period of time.

What do you mean by a "kill switch"?

Do you think that Bitmain has some kind of backdoor in their ASICs which would allow them to seriously damage the equipment? Also, Bitmain controls a lot of hashpower, they could slow down the whole network if they turned off their ASICs right after the difficulty adjustment.

I don't think that Bitmain has a backdoor, I know they do. While it doesn't damage equipment, it does prevent them from mining. This backdoor is in almost all S9s (probably only the first few batches were safe). While you can mitigate it with a hosts file change, I doubt many miners know of this. It's called AntBleed. (https://www.antbleed.com/)


Perhaps it could, but I don't think it would be worth it to attack. There's an extremely high cost for a rental, just to mine some coins that you plan on destroying the value at. POS has it's own problems.

Thank you for bringing it up. It's been a long time since I read article about AntBleed. Hardly anyone mentions it these days. Now, the question is, when are they going to use it? If they wanted to damage Bitcoin in favor of Bitcoin Cash dominance then it would be disastrous for the whole crypto community. Bitcoin is one of the most popular cryptocurrencies. Imagine how satisfied some people would be if it happened. The media would quickly spread FUD. I am looking forward to seeing some other company trying to compete over with Bitmain.

IIRC, the worse thing that can happen is that the miner could've been switched off. They can't be used to attack Bitcoin since they could all be pointed at different pools.

It wouldn't make sense for them to destroy Bitcoin for Bitcoin Cash. They essentially run on the same algorithm. Any attacks on Bitcoin would certainly affect Bitcoin Cash. It creates an impression that Bitmain controls any coin that has SHA256D on it. Bitcoin Gold got 51%'d recently though but it was pretty low cost.

It's a lot more possible for pools to execute a 51% attack than ASIC manufacturers. Pools do not earn all that much from operating one.

Actually, what I meant was that they could slow down the process of solving a block by turning off the miners. Since the difficulty doesn't adjust immediately, it would take it some time for Bitcoin to readjust it (assuming that some people wouldn't fix their ASICs). Bitcoin would be unusable for some time. I was also thinking if it was possible for them to disable fan control and other security measurements which prevent the device from overheating. This could quickly kill the miner.

As you said, it would certainly affect Bitcoin Cash. It is very unlikely that they will try to get rid of the gold mine, but it is unacceptable to see this kind of backdoor.

Eh, that would be the least of our concerns. If 50% of the network turns off, then the block intervals would only increase by 2x which means it would take 10 minutes per block. This wouldn't last all that long; miners could be losing thousands an hour and its unlikely that they are all in the same timezone.
Doubt so. They don't have any motivation to kill the ASICs. They are an ASICs producer and people are going to buy the ASICs from them in the future. They are not going to sabotage themselves by essentially proclaiming to the world that their ASICs are unreliable and no one should buy from them in the future.

The hackers are not going to buy 2 billion hardware to take control of Bitcoin ... But according to the site "51 Crypto" they can control the blockchain (Bitcoin) for 680,000$ per hour :

https://www.crypto51.app

680,000$ is " nothing " ... Taking control just 01:00 would be devastating...

Can you confirm the information of this site?

Um, using 2 billion dollars of USD. If they want to execute an attack, they would have either a participant with sufficiently high hashrate or someone controlling a pool with that much. Once you get the hardware, electricity costs shouldn't be that huge of a concern.

The website assumes that someone has 51% of the hashrate and is willing to lend them out to the attacker at the rate that NiceHash is offering. Plausible, of course. I doubt the total amount that can't profit from mining but still turn their ASICs would collectively have 51% of the network. 1 minute is nowhere near enough to do anything to Bitcoin. You need at least 20 minutes. Reward isn't big either, unless you're telling me that the person isn't requiring at least 3 confirmations.

Well, depends. If someone comes into the game with a chip that can basically reduce the energycost and increase the speeds majorly they would have a bigger chance of finding the blocks and thus with enough funding they could potentially reach 51%.

Look at the mining pools of the world, currently China are ruling this space and we don't know how close the pools are to each other.

ethreum is almost 50% more expensive to attack wow

how can be that ETH is more expansive to attack??

Strange history.....