Cost of 51% Attack $2,270,712,250.38 (rindex.io)

[vit05]

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.



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

[1714839874]

1714839874

[1714839874]

1714839874

[1714839874]

1714839874

[ranochigo]

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.

[ABCbits]

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.

[Ix]

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).

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.

[ranochigo]

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.

[Ix]

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.

[buwaytress]

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.

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.

[DevilOper]

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

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).

[BrewMaster]

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

There is no need to by 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.

[vit05]

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

There is no need to by 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?

[Rath_]

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.

[homeidea11111]

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

[Rath_]

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.

[DevilOper]

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.

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).

[ranochigo]

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.

[BarryBomb]

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

[LtMotioN]

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.

[DevilOper]

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

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'.

[r1s2g3]

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.

[veleten]

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



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