mariusaab on Nostr: Consensys recently published a letter to the SEC stating that Ethereum is more secure ...
Consensys recently published a letter to the SEC stating that Ethereum is more secure than Bitcoin because it would be more expensive to attack (~$34 billion versus ~$20 billion) 💸
Here's why this claim is incorrect:
I find the logic behind Consensys's claim relatively dishonest 🤥 as it relies on the simplistic assumption that having money 💰 is enough to compromise Bitcoin. However, the security of Bitcoin is not measured by the monetary cost of an attack, but by its feasibility ✅
To attack Ethereum, one would need to possess at least 34% of the ETH staked🔒to attack the network, which would equate to about $34 billion, according to Consensys data.$34 billion represents a considerable sum, which few individuals would be able to gather. Moreover, this amount is likely to increase over time if the network gains adoption and if the price of Ether continues to appreciate 📈
This figure still seems underestimated to me, especially if we consider that such purchases would increase the price of ETH concurrently, also increasing the total actual cost of the attack 💹
Considering the hypothesis where it was the United States government 🇺🇸 that decided to launch such an attack, one might wonder why it would not choose to seize the ETH scattered here and there, including those staked on Coinbase, which alone represent 14% of the ETH staked. This would cost significantly less, but it is true that the attack could come from elsewhere 🌎
Regarding Bitcoin, Consensys claims that the cost of an attack is essentially based on the acquisition of computer hardware and the purchase of energy needed to power the machines 🔌
This is true, but the way it is formulated in the letter suggests it would be a simple process, almost reduced to placing an order for a few billion dollars with Bitmain.I think the $20 billion estimate is relatively accurate.
•My quick calculations indicate:An S21 with 200 TH/s costs about $5,000 today 💻
•With about 600 EH/s, it would take 1.53 million S21s to achieve 51% of Bitcoin's hashrate 🔢
•Which equates to about $7.65 billion 💲
Note that this evaluation considers the current cost of the machines. The mass purchase of mining equipment would also increase their price, which would raise the total cost 🔝
Additionally, the buyer would have to obtain these machines before they become obsolete or the hashrate increases too much, otherwise, the cost would be higher.
To this, maintenance fees, electricity costs, transportation of the machines, etc., must also be added ⚙️
However, a crucial detail overlooked by Consensys is the impossibility of acquiring 1.5 million S21s or equivalent models quickly. Manufacturers' order books are already full, and miners wait months or even years to equip their operations ⏳
The idea of quickly acquiring $8 billion worth of mining equipment is unfeasible 🚫
The only way to do this would be for the United States government, to take this example, to manufacture its own mining computers. This would require developing an alternative to market computers and especially the ability to produce 1.5 million machines!
Where would it find the necessary raw materials? What other industry would need to be paused to allow this production? What would be the human, financial, and environmental cost of such an operation? 🤔
Consensys does not seem to have integrated all this into its letter.Moreover, the United States produces 4,500 TWh of electricity annually, while Bitcoin's consumption is estimated between 100 and 170 TWh. The country could theoretically increase its electricity production by 1.5% to cover this 💡
However, the same questions arise: What would be the human, financial, and environmental cost of redirecting between 50 and 85 TWh of electricity to their mining farms? 🌍
In short, the comparison by Consensys doesn't really make sense; it compares one network that relies on a currency to another that relies on energy ⚖️
Even if the estimates are correct, they do not faithfully represent the real cost of a Bitcoin attack
Here's why this claim is incorrect:
I find the logic behind Consensys's claim relatively dishonest 🤥 as it relies on the simplistic assumption that having money 💰 is enough to compromise Bitcoin. However, the security of Bitcoin is not measured by the monetary cost of an attack, but by its feasibility ✅
To attack Ethereum, one would need to possess at least 34% of the ETH staked🔒to attack the network, which would equate to about $34 billion, according to Consensys data.$34 billion represents a considerable sum, which few individuals would be able to gather. Moreover, this amount is likely to increase over time if the network gains adoption and if the price of Ether continues to appreciate 📈
This figure still seems underestimated to me, especially if we consider that such purchases would increase the price of ETH concurrently, also increasing the total actual cost of the attack 💹
Considering the hypothesis where it was the United States government 🇺🇸 that decided to launch such an attack, one might wonder why it would not choose to seize the ETH scattered here and there, including those staked on Coinbase, which alone represent 14% of the ETH staked. This would cost significantly less, but it is true that the attack could come from elsewhere 🌎
Regarding Bitcoin, Consensys claims that the cost of an attack is essentially based on the acquisition of computer hardware and the purchase of energy needed to power the machines 🔌
This is true, but the way it is formulated in the letter suggests it would be a simple process, almost reduced to placing an order for a few billion dollars with Bitmain.I think the $20 billion estimate is relatively accurate.
•My quick calculations indicate:An S21 with 200 TH/s costs about $5,000 today 💻
•With about 600 EH/s, it would take 1.53 million S21s to achieve 51% of Bitcoin's hashrate 🔢
•Which equates to about $7.65 billion 💲
Note that this evaluation considers the current cost of the machines. The mass purchase of mining equipment would also increase their price, which would raise the total cost 🔝
Additionally, the buyer would have to obtain these machines before they become obsolete or the hashrate increases too much, otherwise, the cost would be higher.
To this, maintenance fees, electricity costs, transportation of the machines, etc., must also be added ⚙️
However, a crucial detail overlooked by Consensys is the impossibility of acquiring 1.5 million S21s or equivalent models quickly. Manufacturers' order books are already full, and miners wait months or even years to equip their operations ⏳
The idea of quickly acquiring $8 billion worth of mining equipment is unfeasible 🚫
The only way to do this would be for the United States government, to take this example, to manufacture its own mining computers. This would require developing an alternative to market computers and especially the ability to produce 1.5 million machines!
Where would it find the necessary raw materials? What other industry would need to be paused to allow this production? What would be the human, financial, and environmental cost of such an operation? 🤔
Consensys does not seem to have integrated all this into its letter.Moreover, the United States produces 4,500 TWh of electricity annually, while Bitcoin's consumption is estimated between 100 and 170 TWh. The country could theoretically increase its electricity production by 1.5% to cover this 💡
However, the same questions arise: What would be the human, financial, and environmental cost of redirecting between 50 and 85 TWh of electricity to their mining farms? 🌍
In short, the comparison by Consensys doesn't really make sense; it compares one network that relies on a currency to another that relies on energy ⚖️
Even if the estimates are correct, they do not faithfully represent the real cost of a Bitcoin attack