how is bitcoin not a massive waste of energy?

You learn about Bitcoin, it seems very futuristic, the idea of just currency with a fixed rate of inflation… what is it, roughly 50 bitcoins are born every 10 minutes. What concerns me is this Mining process.

As many people as want to run these Mining servers, they form the backbone of the system, and it is my understanding that they act as a sort of distributed clearing house and generator of new bitcoins in the system.

How Bitcoins are Generated

To generate new Bitcoins, all of these servers are constantly attempting to solve a cryptographic problem (of a known difficulty). Say there are 10 mining servers out there, the system automatically adjusts the difficulty of the problem such that one of these ten servers is going to generate a Bitcoin block (50 bitcoins) once every 10 minutes. If there are a million people running mining servers, the system bumps up the complexity in order to make sure that only one of these servers finds a solution.

In other words, there are thousands (maybe one day millions?) of servers out there being asked to churn on busywork for Bitcoin. This isn’t just any busywork, people are throwing massive GPU power at this stuff to generate Bitcoin. The people who run them are, literally, banking on the idea that if they run more Bitcoin servers they are more likely to land a new Bitcoin block. This is like creating money out of thin air…. or is it?

Bitcoin’s Carbon Footprint

Think of this difficulty adjustment as bumping up the carbon footprint of Bitcoin. The more people that run mining servers in this system, the higher the difficulty of the solution, the larger the footprint in terms of energy used. So much energy is being used that the community has started to compare efficiency ratings of various rigs and Bitcoin miners have been raided because the energy profile resembles that of an indoor grow operation.

On the one hand, I understand that Bitcoin mining is the incentive that the network uses to encourage people to setup the infrastructure that makes Bitcoin possible. On the other hand, doesn’t anyone see how, if and when Bitcoin becomes useful, it starts to consume more and more energy.

Since most of electricity is from coal, someone should relate the current difficulty factor for Bitcoin mining to “number of metric tons of coal produced”. As of March 2011, people were talking about the power of Bitcoin’s network as being in the ~5 PetaFLOPs range. If we look at the cost of a FLOP from Wikipedia it lists 1 GFLOP at around 1.29 watts.

You see those power plants over there, they generate 50 bitcoins every ten minutes…

If Bitcoin is running ~5 PetaFLOPS and the energy of a GFLOP is 1 watt, this means that we’re throwing Bitcoin at least 5 MW just so that one mining server somewhere in the world can generate one Bitcoin block every 10 minutes. If the amount of mining hardware increases by a factor of 10x, this could go up to 50MW – still for 50 bitcoins every 10 minutes for one lucky server in the network.

Doesn’t this just seem like an awful waste of both computing power and energy? The network is still small, but multiply that by a factor of 10x and we’re devoting a power plant to Bitcoin.

I would appreciate someone pointing out some flaws in this logic because I do find Bitcoin interesting. I’m not anti-Bitcoin, it is just that the energy consumption question was bugging me during Don Park’s Open Source Bridge Conference presentation. The conference was held in a Green building in Portland, we’re looking out the window at a building with wind turbines on it. The conference is trying to recycle, everyone’s aware of the imperative to “do the right thing”. When Don mentioned efficiency as a goal, everyone nods “great”, but isn’t the question: “Why base Bitcoin creation on energy at all?”