Online Habits Killing The Planet - Dispatches

[kalniel]

I'm far more interested in the energy levels of ETH.

Sorry my mindreading failed me, when you asked about bitcoin transactions I thought you meant bitcoin transactions

[wazzickle]

Come on, are you really going to tell me that you read me saying 'bitcoin' and didn't understand that I meant 'eth'?

[DanceswithUnix]

That would be useful except for I don't have a tesla, I drive a mazda 3 that gets about 40-45 mpg. Yes, I know they're very hard to compare, but I've seen such comparisons before, so I know it's possible.

I'm far more interested in the energy levels of ETH.

That is interesting, we see lots of talk about Bitcoin but none about the altcoins, and Eth has to be by far the worst energy consumer of the lot (from pure size of adoption compared to the others, coupled with using proof of work).

I shall throw out an estimate, and see what holes people pick in it...

Eth has a total hash rate today of 575 THash/s : https://etherscan.io/chart/hashrate

We can put some bounds on energy consumption for that.
Now I suspect most of the mining will be done on 7nm graphics cards, the 5700XT and 3060 Ti seemingly the ones to go for. So a quick Google tells me you can expect 60MH/s for 120W on a 3060 Ti, so if everyone was using those cards (yes I know they aren't, this is getting a feel for the power usage you can't get a precise number) then you are looking at

(575000000/60)*120 = 1.15GW

I will admit that is a bit better than I was expecting, the world's Eth consumption is currently about the output of Sizewell B, our largest nuclear power plant in the UK. Or, as I type this, the entire solar power plant output of the UK: https://grid.iamkate.com/
As an aside, I suspect the 575000000/60 = 9.6 million is a rough indicator of how many GPUs are mining (just Eth) at any one moment.

That gives a median value, and I suspect isn't that far out. Now, the possible spread of how far out that could be.

There will be some people running older 12nm & 14nm graphics cards, and a quick Google of the GTX 2080 and RX580 it seems 100W gets you 25MH/s, so if all that mining was on older cards you could about double that figure to about 2.2GW

If everyone was using ASIC miners, then those things are 3GH/s at 800W, so we are looking at (575000000/3000)*800 = 153 MW, which is a lot but clearly nowhere near as bad. That's why I am saddened by the Crypto community's resistance to ASICs.

So the total energy consumption of just the GPUs in the eth network is between 2GW and 0.15GW, but probably about 1GW assuming most users are on 7nm GPUs and the 14nm GPUs and ASIC miners sort of balance each other out. Now the CPU and motherboards driving those will consume a bit, and the PSUs will have some overhead so you can add maybe 10% onto that given mining rigs are run with an eye on efficiency using high efficiency PSUs, low end Pentium class CPUs and one stick of ram.

Now the bit that gets sickening... That ~ 1.2GW is powering the world computer to do about 16 transactions per second: https://blockchair.com/ethereum/char...ons-per-second.

16 transactions/s is about peak trading for Tesco. Traditional payments, I can get better than that out of a single raspberry pi.

[kompukare]

And using DanceswithUnix's figures and the hash explorer gives this graph:

So if the total hashrate is nearly 10 million cards, it seems most of that was added recently - although the previous dips might imply some dual-usage cards in that when things boom gamers might mine when not gaming.
The rise coincides with the release of Ampere since then hash has increased from 262 to 573. So around half of those 10 million cards were added recently - 5 million cards or so.
Not really going anywhere with this, however:

1. The March 21 Steam survey implied that around 3.5 million 'gamers' (or people who install Steam) have Ampere cards
2. This gen Nvidia mines better than AMD
3. AMD's RDNA2 is almost totally absent on Steam.
4. Consoles have sold over 8 million by now

Not sure what this all implies, but demand is crazy even without mining. From the hash rates it is possible that Ampere has sold over 5 million or more with 2 million or so to pure miners.

[DanceswithUnix]

[*]This gen Nvidia mines better than AMD

Note that the 5700X was really good at mining (one Youtuber bought 70 of them at the start of the current bubble). The latest 6000 series cards with the narrower ram and big cache act as a nice hardware mining limiter, so no bypassing that They only have to be profitable vs their power usage though, rose tinted mining glasses will do the rest to seal the AMD purchase.

Those noticeable dips in the graph are probably rigs being switched over to some other coin as it becomes more profitable, then being switched back again.

[wazzickle]

Those noticeable dips in the graph are probably rigs being switched over to some other coin as it becomes more profitable, then being switched back again.

Why wouldn't they be explainable by virtue of price, i.e. wouldn't the demand for cards track with profitability of mining?

[DanceswithUnix]

Why wouldn't they be explainable by virtue of price, i.e. wouldn't the demand for cards track with profitability of mining?

Not from what I have seen. Those rose tinted mining glasses seem to make all cards look profitable regardless of the price of the cards, it is just whether you get your money back in two months or in 12 months. OFC, if you *really believe* like a Disney Princess, then "when crypto takes over and the value goes to the moon" even running at a loss now leads to profit later. Not my idea of sane risk, but there you go.

Now the Eth gas price has dropped back to something sane recently. That will really hit Eth miners in the wallet, with an outlook of mining rewards getting lower in future. But with current Eth prices where they are (and they could well sneak up further) I don't think any of that matter.s

[TheAnimus]

To explain the 'take the power back' line, well, that's something slightly different from what I mean. The banking system is corrupt as hell. It is meant to be something that provides liquidity to the system. What it actually does (amongst many other things, some good, some bad) is transfer wealth and power from the many to the few. It is a key cog in the system - perhaps, if you take the analogy of the car, it is the actual wheels. You can see the way our society is moving away from cash - for understandable reasons, governments want to be able to keep tabs on their citizens. The libertarian stance on crypto is that it's stopping that process. I'm no libertarian, mind, what I class as government overreach is different from what the average crypto permabull does, but nonetheless, I see the appeal. Any movement away from central bank control of our funds is a good one. You can call it cringeworthy if you like - that's a personal opinion - but unfounded it is not. There is lots of legitimate political philosophy behind cryptocurrency.

So by taking power back, you mean shuffling the deck, the vast, vast majority have no chance of getting into the asset bubble, because that's what it is due to not having enough "coins" per head of global population.

There is no way a disenfranchised person in a 3rd world country is benefitting.

As for the tabs on their citizens? Cash is far harder to track than crypto (oh and cash works at scale).

Why do you not want a central bank involved, their reason for being is to provide stability. This is why we regulate banks because we learnt that we had to in the 19th century. Hence why crypto is so unstable, unsafe and completely useless.

[DanceswithUnix]

Another datapoint on the whole mining thing.

This should be a significant sample of home mining. Not the big players who self host everything, don't bother with pools etc:

https://hiveos.farm/statistics/

The modal card is an AMD 580/570, so I guess people really don't care about energy use and I was underestimating a bit.

[wazzickle]

A good article on the business of mining and why the attacks on it are at least unfair, at most politically motivated.

https://news.bitcoin.com/bitcoin-ene...anking-system/

[wazzickle]

So by taking power back, you mean shuffling the deck, the vast, vast majority have no chance of getting into the asset bubble, because that's what it is due to not having enough "coins" per head of global population.

There is no way a disenfranchised person in a 3rd world country is benefitting.

As for the tabs on their citizens? Cash is far harder to track than crypto (oh and cash works at scale).

Why do you not want a central bank involved, their reason for being is to provide stability. This is why we regulate banks because we learnt that we had to in the 19th century. Hence why crypto is so unstable, unsafe and completely useless.

Yes, shuffling the deck would be a fairer distribution. The fact that it wouldn't benefit poor citizens in the third world has as much relevancy as the fact that you can't save all the people tortured to death by the Spanish Inquisition by virtue of driving less.

[watercooled]

Oh of course. Bitcoin.com is providing an entirely impartial and unbiased view of the matter, and not pro-Bitcoin at all.

Having just viewed the article I can already see it falls into many of the same fallacies I see everywhere else, some of which I have covered in this thread. Not least of which are the 'but it uses renewable energy' nonsense and comparing an established financial system which serves the world to an almost totally insignificant system barely capable of handling a vanishingly small fraction of a single country's transactions, yet managing to register on a global scale in terms of resource consumption. Oh and it throws in the 'but but other stuff wastes electricity too' argument for good measure. And lets gloss over the fact it tries to dismiss its critics by resorting to ad hominems; it manages to get the terms woke, cancel culture, emotional and virtue signalling into one paragraph! I would expect more professional writing in an article making out to be objective.

Dismissing criticism as 'politically motivated' is just a convenient way of trying to dismiss said criticism without actually answering it. Why do you think the criticism is wrong?

[DanceswithUnix]

A good article on the business of mining and why the attacks on it are at least unfair, at most politically motivated.

https://news.bitcoin.com/bitcoin-ene...anking-system/

It isn't that proof of work uses power, it is that PoW is being so badly mishandled in bitcoin.

Here's a simple thought experiment...

Let's say you limit the network to 10000 machines (either hard limit or some other mechanism), and you set up your crypto in a way such that their compute power isn't significant. That would push people into using the most efficient platform possible, and you end up with a network of raspberry Pi's with a total consumption on par with a kettle or two running all the time. A few hippies would complain because there is always someone, but basically you are golden.

Would that be useful? Yes, you could happily run the Bitcoin network on 100 Raspberry Pi machines and AFAICS the transaction time would be the same as it is right now.

So what went wrong? Well it is a race where each miner wants the highest percentage of the network they can get. The network size is unbounded, and the performance per node is unbounded. So arguing what the precise power consumption is misses the point, as that is always a starting point for the next massive phase of growth. At the end of that growth, the bitcoin network will be consuming more power with no benefit to any of the users. The growth also favours those who can throw the most money at it, so you end up with Bitcoin network being essentially owned by a few rich people. It is worth them doing that because they can make mining profits, taking transaction fees from users.

So a few rich people are getting richer by taking a fee from the masses. The only difference between this and the conventional banking system is that the conventional banking system is faster and cheaper, and it can largely do that because it isn't peeing energy up the wall.

If you've actually read this far then here's the political bit for you. Find another coin, and evangelise that. Storj pays coins for performing a task people need, Stellar allows payments in a cheap and energy efficient manner. There are plenty of alternatives out there, and we should be pushing those until the cancer that is Bitcoin dies.

[wazzickle]

Oh of course. Bitcoin.com is providing an entirely impartial and unbiased view of the matter, and not pro-Bitcoin at all.

Having just viewed the article I can already see it falls into many of the same fallacies I see everywhere else, some of which I have covered in this thread. Not least of which are the 'but it uses renewable energy' nonsense and comparing an established financial system which serves the world to an almost totally insignificant system barely capable of handling a vanishingly small fraction of a single country's transactions, yet managing to register on a global scale in terms of resource consumption. Oh and it throws in the 'but but other stuff wastes electricity too' argument for good measure. And lets gloss over the fact it tries to dismiss its critics by resorting to ad hominems; it manages to get the terms woke, cancel culture, emotional and virtue signalling into one paragraph! I would expect more professional writing in an article making out to be objective.

Dismissing criticism as 'politically motivated' is just a convenient way of trying to dismiss said criticism without actually answering it. Why do you think the criticism is wrong?

I've already gone over all the reasons I think the criticism is wrong, we'd be going in circles if you asked me to repeat myself. I'm not claiming that bitcoin.com would be impartial, but rather that the arguments and evidence they've collected could be useful in the context of this. Consider it like me and you the judges asking the defence lawyers to step forward and present their evidence.

Again, I'm not dismissing anything. Being politically motivated is ok. It just means you have politics.

I found the article to be reasonably well-written. If you think it's not, you could see that as a lens on our differing standards on writing, or you could see it as a lens on our differing levels of emotional reaction to the content.

[wazzickle]

It isn't that proof of work uses power, it is that PoW is being so badly mishandled in bitcoin.

Here's a simple thought experiment...

Let's say you limit the network to 10000 machines (either hard limit or some other mechanism), and you set up your crypto in a way such that their compute power isn't significant. That would push people into using the most efficient platform possible, and you end up with a network of raspberry Pi's with a total consumption on par with a kettle or two running all the time. A few hippies would complain because there is always someone, but basically you are golden.

Would that be useful? Yes, you could happily run the Bitcoin network on 100 Raspberry Pi machines and AFAICS the transaction time would be the same as it is right now.

So what went wrong? Well it is a race where each miner wants the highest percentage of the network they can get. The network size is unbounded, and the performance per node is unbounded. So arguing what the precise power consumption is misses the point, as that is always a starting point for the next massive phase of growth. At the end of that growth, the bitcoin network will be consuming more power with no benefit to any of the users. The growth also favours those who can throw the most money at it, so you end up with Bitcoin network being essentially owned by a few rich people. It is worth them doing that because they can make mining profits, taking transaction fees from users.

So a few rich people are getting richer by taking a fee from the masses. The only difference between this and the conventional banking system is that the conventional banking system is faster and cheaper, and it can largely do that because it isn't peeing energy up the wall.

If you've actually read this far then here's the political bit for you. Find another coin, and evangelise that. Storj pays coins for performing a task people need, Stellar allows payments in a cheap and energy efficient manner. There are plenty of alternatives out there, and we should be pushing those until the cancer that is Bitcoin dies.

Man, people feel alarmingly strong about this, and I'm not quite clear why.

I'm not evangelising bitcoin, should you think that I am. As long as any other cryptos are around, bitcoin won't die, and it's certainly not cancer.

I'm also not claiming it as pure and the saviour of mankind or the banking system. I'm just saying it's less bad, for a number of reasons, and the downsides to it are being overplayed by people with vested interests, which is no surprise, as no-one likes it when their hegemonies are challenged. I'm not even criticizing the banking system or mainstream media for making themselves party to this attack, it's what I would do. But it's still just that. Not relevant in the wider context of the actual villains in the piece: the american military and the other gigantic international corporations who laugh at us trying to use less plastic while the damage they do to the environment dwarfs any possible savings we could make by revolutionising our entire lifestyles. It's a distraction.

[watercooled]

I've already gone over all the reasons I think the criticism is wrong, we'd be going in circles if you asked me to repeat myself. I'm not claiming that bitcoin.com would be impartial, but rather that the arguments and evidence they've collected could be useful in the context of this. Consider it like me and you the judges asking the defence lawyers to step forward and present their evidence.

Again, I'm not dismissing anything. Being politically motivated is ok. It just means you have politics.

I found the article to be reasonably well-written. If you think it's not, you could see that as a lens on our differing standards on writing, or you could see it as a lens on our differing levels of emotional reaction to the content.

To be honest I didn't actually notice I was replying to the same person until after I initially posted that most recent reply. I did consider re-wording it but decided to leave it as-is. I wasn't sure why the person I originally had the debate with had just presented the exact same discredited arguments again TBH.

What I am reading in that article is not evidence, it is opinion and rhetoric, and frankly you did a better job of arguing your point without it. There is nothing in that article that we didn't already know, and they table the same stale arguments we've seen time and time again. As I've said before, making the same demonstrably false claims repeatedly doesn't actually make them true. Nor does being published on the bitcoin website.

I'm not sure I'd consider the arguments I have presented on the matter as 'political'. Perhaps in the sense that totally unnecessary wastage, pollution, market disruption and misinformation irks me, but not in the anti-crypto strawman sense that I often see people making out. As Dances said very well above, what's the issue with a more intelligently designed system that doesn't encourage needless waste, and actually does the job better (or even remotely close to as-good would be a start) than the system it claims to want to replace or supplement? Better alternatives already exist, but adoption and hype seem to revolve largely around people wanting to profit from mining at any cost.

You could separate the fundamentally anti-crypto people from those who are just pragmatically against certain aspects or implementations by considering that, rather than conflating any criticism of Bitcoin as anti-crypto propaganda.

Man, people feel alarmingly strong about this, and I'm not quite clear why.

Some people hate to see unnecessary and increasing wastage at a time when changes and sacrifices are being made to try to offset the impact we are having on our environment. That's one perfectly valid reason, to start.

and it's certainly not cancer.

I'm playing devil's advocate a bit on this one, but I don't see it as unfair to describe it as such. Mining is largely driven by greed and availability of resources. I have used the term parasitic for much the same reason.

I'm also not claiming it as pure and the saviour of mankind or the banking system. I'm just saying it's less bad, for a number of reasons, and the downsides to it are being overplayed by people with vested interests, which is no surprise, as no-one likes it when their hegemonies are challenged. I'm not even criticizing the banking system or mainstream media for making themselves party to this attack, it's what I would do. But it's still just that. Not relevant in the wider context of the actual villains in the piece: the american military and the other gigantic international corporations who laugh at us trying to use less plastic while the damage they do to the environment dwarfs any possible savings we could make by revolutionising our entire lifestyles. It's a distraction.

I'm curious what you think our/my vested interests are?