In this article we will explain what a 51% attack is, how much energy it requires and what damage it would mean to Bitcoin.
What is a 51% attack?
A 51% attack occurs when a group of Bitcoin miners controls more than 50% of the Bitcoin network's computing power. With this majority, they could manipulate the network by reversing transactions or double spending. This could lead to loss of trust, loss of value and other negative effects.
Are mining pools capable of this?
Some miners join together to form so-called mining pools, which allows them to pool their computing power. This increases the probability of finding a valid block and receiving the block reward. This is then divided among the members of the mining pool, depending on how much computing power an individual member has made available to the pool.
Incentive system
To understand how likely a 51% attack really is, it is necessary to understand the incentive system of Bitcoin miners. If the pools, as a group of miners, try to add an invalid block to the blockchain, it would be rejected by the network (full nodes) and the miner would therefore forego the corresponding block reward. Each miner therefore has an economic incentive to act honestly in order to receive the block reward. This is the only way a miner can cover their business costs in the long term.
Mining offline
There are also so-called double-spending attacks, in which malicious network participants can spend their Bitcoin multiple times. Basically not possible for mining pools. In order to secretly mine a “fake blockchain,” the pool operator would have to take it “offline” and publish the new chain at a later date to overwrite the old one. However, if the operator goes offline, the participants in the pool would immediately jump out and change it. Since this can literally be done in just a few minutes, it can be stated that pool operators actually have no power and should rather be seen as service providers. Malicious miners in general would suffer financial damage anyway, because the mining hardware purchased for the attack would also quickly lose most of its value. Due to the enormous capital costs, there is no economic incentive for a 51% attack. Every network participant would suffer financial damage. Only an attacker who attempts to attack the network with a self-destructive motive could become a threat to the network. However, this attack would involve enormous amounts of economic resources. However, this is unrealistic for mining pools, as it cannot be assumed that all pool participants want to harm the network. Rather the opposite.
Example Ghash.io
In 2014, Ghash.io controlled more than 50% of the hashrate. The pool's miners could therefore have carried out a 51% attack. Even back then, this situation unsettled the Bitcoin community. But the economic incentives prevailed. Ghash.io released a statement asking miners to leave the pool so that the pool's share falls below 40%. As mentioned above, a big advantage of Bitcoin miners is their flexibility. To belong to a mining pool, miners simply have to enter the so-called stratum address into the mining software and deposit a wallet for payouts. The miners complied with the pool operator's request and moved to other pools. A year later, Ghash.io's market share fell to 2%.
Stratum V2
However, the Bitcoin community is aware of the potential dangers of centralizing mining pools, which is why decentralization is already underway.
The most promising project is Stratum V2, a revised version of the current mining protocol, which was presented at the Bitcoin Conference 2019. With Stratum V2, miners in a pool can decide for themselves which transactions should be included in the block. This is currently the responsibility of the pool operators and thus reduces the dependency on the mining pools. Although block generation is still in the hands of the mining pools, adopting Stratum V2 would be an important step towards decentralizing the mining pools.
How many miners for 51%?
To successfully carry out a 51% attack, a group of attackers requires a significant amount of computing power. The exact amount depends on the current hashrate of the Bitcoin network, which is constantly changing as new Bitcoin miners join and old ones leave. Currently (March 2024) the hash rate is approximately 600 exahashes per second, i.e. 600,000,000,000,000,000,000 hashes per second worldwide, which all Bitcoin miners put together. One of the best and most popular devices for mining Bitcoin is the Antminer S19 Pro from Bitmain. The miner has a hash rate of 110 TH/s (terahashes per second) and consumes 3,250 watts of power. If we now want to know how many of these S19 Pro miners we need to get 51% of the hashrate of the Bitcoin network, we multiply the 600 exahashes by 51% and get 306 exahashes. If we now divide this by the 110 terahashes of a device, we get the number of S19 Pro:
2,781,818 devices
How many nuclear power plants?
How much electricity do we need? To do this, we multiply the number of devices by the 3,250 watts of a device:
9,040,908,500 watts. How many nuclear power plants would we need for this? The two reactor blocks of the Chinese Taishan nuclear power plant currently have the largest net output in the world at 1,660 MW (see Wikipedia). 1 megawatt equals 1,000,000 watts. Accordingly, 1,660 MW is 1,660,000,000 watts that the world's most powerful nuclear power plant, Taishan, can generate. Now we divide the 9,040,908,500 watts required for 51% of the hashrate by the 1,660,000,000 watts of a Taishan nuclear power plant. Result? For 51% of Bitcoin's hashrate, we need 5.44 Taishan nuclear power plants, whose electricity should not be used for anything other than Bitcoin miners. Not only would this project have to take place in secret, but the success would also only be short-lived. Right from the first block by the attacker, every user in the network would be aware through the mining data in the block that 51% of the hashing power was in the hands of an attacker. No further block would now be able to be added to the blockchain by the attacker through the nodes. The network would split and the attacker would automatically leave the network.
What if the attacker knew how to hide the attack?
Even if the attacker managed to ensure that not all nodes rejected his blocks directly, the attacker would have a significant problem after 2016 blocks at the latest, since the so-called difficulty adjustment takes place in the network every two weeks, which is determined by an algorithm that increases the difficulty of the block Mining is always set so that a new block is only found every ten minutes on average. Since the attacker has invested a massive amount of new energy into the network, blocks are now found much more frequently, which would lead to a massive increase in difficulty. This in turn would ensure that the time interval for finding blocks for the attacker would increase significantly.
At this point it is up to each person to decide how likely such a scenario is.
Your Bitcoin miner for home
