[loverofthings]

Compared to heating and cooling, and a bunch of other stuff it is negligible.

I guess you could complain about the usefulness of mining.

Well, people do a bunch of useless stuff that has magnitudes bigger carbon footprint than mining.

Let's not optimize things that are insignificant.

[pavlov]

It's not insignificant. Bitcoin mining already consumes as much energy annually as Denmark – that means the entire country, heating, cooling, industry, everything:

https://arstechnica.com/tech-policy/2017/12/bitcoins-insane-...

[ddebernardy]

Exactly. What more, it "only" consumes as much energy annually as a whole country because the bulk of the mining is currently done using specialized hardware.

If websites begin to crowdsource mining on their visitors' PCs using JavaScript, the energy consumption will be even higher.

[loverofthings]

In the grand scheme of things (context being the carbon footprint of humanity) it is insignificant, just like Denmark's carbon footprint is.

Let's optimize the real stuff, not some trivialities (like the Christmas lights in the USA pointed out as a reply).

There's 24/7 AC in USA and China in the masses. There's Brazilian and Argentinian agriculture removing rainforests to grow food for the USA or EU that grows cattle. (EU takes unbelievable amounts of rainforest grain/legumes to feed its livestock, an absolute travesty and hipocrisy)

The footprint of these developments is insanely large, yet somehow coin mining is an issue.

We could then jump to optimize the Internet infrastructure too, despite being 1-2% of CO2 equivalent footprint.

Yes, I understand coin mining seems a bit useless but I'm pretty sure heating and cooling is wasted in magnitudes more amounts, being applied in badly designed buildings, or cooling the streets of Las Vegas.

[saint_fiasco]

I don't have a good intuition for that kind of heterogeneous comparison.

Can it be compared with other Internet services? How much electricity does Facebook consume, for example? What about internet advertisements?

[wor3q]

It's 5 times what USA use on christmas lights: https://phys.org/news/2015-12-christmas-energy-entire-countr...

Or 3,5 times what Amazon US-East cluster consumes: https://datacenterfrontier.com/amazon-approaches-1-gigawatt-...

[pavlov]

That's a really good question. I don't have the answer but I'd like to know too.

[wor3q]

Denmark has 5M people and average temps between 0 and 17C. That means little heating and no ac.

32tWh/yr means roughly 3,6GW, which is output of 1-2 average nuclear power plants.

[pavlov]

The cost of building a nuclear reactor in the West is in the ballpark of $10 billion these days. It's not a trivial expense.

If we're going to soon need dozens of nuclear reactors to power cryptocurrencies, that's worth discussing at least. You can't brush it away as something insignificant in the grand scheme of things.

[loverofthings]

No one is going to build nuclear reactors in the West for cryptocurrency mining. Just like no one is building nuclear reactors for powering christmas lights or Amazon servers, or Internet servers.

https://en.wikipedia.org/wiki/Nuclear_power_phase-out

Reactors are being built primarily for heating and cooling, and eventually vehicle charging.

[wor3q]

Last months bitcoin trade volume was $273 billion.

Let's say a reactor works for 10 years.

That would give $32,760 billion in trade, and reactor cost would be 0.833% of that.

Not a big fee if you ask me.

Of course, if cryptocurrencies would scale so they become significant between other trades, we can't just built thousands of nuclear reactors. This is where proof of stake will come into play.