I always find these breakdowns to be a little bit disingenuous. Like, you could do this same analysis on the whole email system, or on the whole world wide banking system, including ATMs, or on the energy usage of all DNS queries or even on global ActivityPub activity, not to mention shopping on Amazon or browsing Facebook. People DO do these kinds of breakdowns on generative AI, for exactly the same reasons, and reach the same kinds of conclusions.
Having a global computer network is INCREDIBLY energy intensive, with a massive carbon footprint. It’s not shocking that a given application of that network is energy intensive, with a massive carbon footprint. These kinds of analysis are put together by people who already don’t like cryptocurrencies (for all kinds of reasons both valid and ridiculous) who then go cherry picking MORE reasons not to like them.
A better comparison would be energy utilized per user, in which case the energy requirements for Bitcoin are miles and miles ahead of what the average person produces using a computer in the same amount of time. Even a gamer, playing 4k 120fps ray traced games 12 hours a day would use a fraction of the energy of someone mining bitcoin.
Exactly, if we do a back of the napkin calculation:
Bitcoin
Users
There are 200 million bitcoin wallets, let’s be generous and say those are all owned by unique individuals.
Total energy consumption
Bitcoin used about 114 TWh in 2021[1]
Bitcoin currently uses about 150 TWh annually
Energy consumption per user
150 TWh / year ————————— = 0,75 TWh / user / year 200 million usersBanking system
Users
There are over 8 billion people on the planet today, let’s assume 4 billion of them have access to the global banking system.
Total energy consumption
The global banking system used an estimated 264 TWh in 2021[1]
If we assume the same consumption increase rate for banking, that’s about 348 TWh/year currently.
Energy consumption per user
348 TWh / year ————————— = 0,087 TWh / user / year 4.000 million usersWith these numbers, bitcoin uses almost 10x the energy per user annually.
There are of course a myriad of things one can argue over whether it makes a fair comparison, none of which I feel like arguing, since this is just a really simple estimate with a lot of assumptions.
1: I used the numbers in this article uncritically, if you have better numbers you can run your own calculations.
