The heat death of the (bitcoin) universe

Wednesday 22 June 2016

Bitcoin mining is energy intensive. Just how problematic is that, long term?

Bitcoin’s proof-of-work is a remarkable system. Ten years ago it wasn’t clear that the double-spend problem - the issue that digital information can readily be copied, and therefore transactions easily duplicated - could be solved. Satoshi Nakamoto did just that, thus enabling for the first time truly peer-to-peer transactions that require no third party to authorise them.

Bitcoin mining works by, essentially, requiring computers to generate random numbers in the form of hashes: a repetitive but simple mathematical process that is computationally expensive. These hashes have to fit the criteria set by the bitcoin protocol; that is a moving target depending on how much total processing power is being thrown at the network. The computer that generates the winning hash gets a block of bitcoins, currently 25, as well as validating the most recent set of transactions from users.

Hard work

Mining bitcoins is hard work - and it's only going to get harder

Hard work

It’s a brilliant way to address the double-spend problem, but it requires a lot of energy. As the UK government’s ‘Beyond Blockchain’ report notes, ‘The mining analogy is apt because the process of mining Bitcoin is energy intensive as it requires very large computing power. It has been estimated that the energy requirements to run Bitcoin are in excess of 1GW and may be comparable to the electricity usage of Ireland.’

So here’s the problem. Energy costs money. Even if you’re using renewable power, there are costs involved. The relationship between hashrate and bitcoin price is subject to endless debate, but the reality is that if the price of bitcoin is low, miners find it uneconomical to run their rigs; hashrate has dropped at points in recent months as the price fell to medium-term lows.

Supply and demand

Conversely, if bitcoin’s price rises, so will the amount of hashing power, since new miners will find it economical to come on board. Supply and demand.

You can make all kinds of cases that miners might be holding coins, taking a loss on energy costs in the hope of future gains, and so on. But the fact is that right now, the market for both BTC and hashrate tells us that miners are prepared to keep their rigs running at this price. In fact, hashrate is steadily rising.

Multiply the value of a bitcoin by 10x, and you logically multiply the energy costs by 10x, ceteris paribus. At the time of writing, bitcoin has a price of $660 and hashrate stands at 1.585 GH/s. Assuming widespread adoption and $50k bitcoins, you’re looking at potentially well over 100 GW, or around 1% of all human energy use.

Now, this doesn’t have to be a dealbreaker. The blockchain in its current form isn’t designed to scale like this, and it’s reasonable to expect other systems such as sidechains to come to bear for many kinds of transactions. Proof-of-stake platforms are far less energy intensive, and are also suited to certain applications. All it means is that, unless we crack the nut of nuclear fusion sooner than expected, bitcoin itself and in its current form is unlikely to become a currency of mass use. But there’s plenty of time to solve that problem, and plenty of more pressing issues to deal with first. It's just one for the devs to add to the list.

comments powered by Disqus