Good way to spin having only 4GB vram into an 'anti-mining feature' I guess.
Of course, 4GB is typically considered the bare minimum for modern 1080p (let alone higher) gaming and this is just a budget card which was definitely not limited with any mining considerations.
The fact that the entry-level card can run its clock 30% faster than the same line's top-end offering supports the story pretty solidly, IMO. Not supporting 8 GB of vram can't have given them enough breathing room to casually run the clock that fast, so this has to have been decided way up front.
Like the article says, it remains to be seen whether this will also make the card bad at gaming, but I think that AMD at least had a coherent plan here.
RDNA is far more efficient at gaming. GCN was compute-focused (higher TFLOPs, but weaker in practice). Case in point, micro-benchmarks show that VRAM latency is ~100 nanoseconds on RDNA, but ~300 nanoseconds on GCN.
RX 6500 XT is aimed at roughly the same specs as the RX480, but should perform slightly faster in practice, along with Raytracing support, Infinity Cache and therefore faster VRAM latency (probably faster fill-rates and other microbenchmarks important to gaming), all at 107W instead of 150W.
---------
As such, Rx 480 would be a better mining card, while 6500 XT would be a worse mining card, but better video game card.
Its still stupid for AMD to market their cards like this, but they're not lying... (and whatever argument marketing wants to make... well... that's their job. It all looks stupid to me but I can't honestly call them out on it since it is true...)
-------
Turns out that video-gamers who want more uniformity and less jitter between frames want higher-clock speeds (notice the 2000+ MHz clock speed on 6500 XT) rather than raw TFLOPs (Rx 480 ran a slow clock of ~1000MHz but had many more compute units / SIMD units computing in parallel).
Width of GPU leads to more variance, because its harder to load-balance your parallelism. If you have a narrower GPU (fewer compute units) at higher speeds, its easier to reach more consistent frame times / less jitter.
Further, given the prices of RX480 and modern video cards, even if they were on par, if you can actually GET RX6500XT for $200 today, then:
1. It sucks from big picture perspective, as we are "back where we were 5 years ago"
but
2. It's great from pragmatic perspective, as we are otherwise today in "Wish I could afford a card from 5 years ago" / "there's no such thing as a budget GPU" stage.
If we look at the benchmarks of the 5500XT vs the RX480 I expect it to maybe be 5-8% faster if at all. AMD basically released this card 3 times now but this time they scrubbed the whole encoding feature set.
> AMD basically released this card 3 times now but this time they scrubbed the whole encoding feature set.
Do you not see the difference between GCN, RDNA, and RDNA 2?
At a minimum, this RDNA 2 6500 XT has raytracing cores that the two previous versions do not have. I'd assume that most video gamers care more about the raytracing feature (especially as more and more games are using it) than H265 encoding.
Remember: Raytracing instructions were invented on RDNA2. The 5500 doesn't have it, and the 480 doesn't have it.
-------
I mean, seriously. Would you rather have the existence of this card? Or the non-existence? If you want something faster, that's why the 6800 exists. This 6500 XT is for people who want a barely-gaming ready card at the cheapest possible price. If that's not your use case, then don't buy the card.
But every now and then, I build a computer for a child who doesn't play many video games, or for some uncle/aunt who doesn't know much about computers. I never like the idea of going iGPU only, because these people inevitably find at least one weird game that they like to play. So I like to put in a $100 to $200-class card in there. Not expensive enough to break the budget, but high enough that it'd perform decently on 10-year-old games and maybe a few indie-games.
Are you suggesting that you could use Raytracing at all in modern games on a low end card like the RX 6500 XT? If you would use it you would lose the marginal performance increase of the updated architecture. Also content creation is a big point for younger generation if it is livestreaming, recording clips or general gameplay. This will both not be doable with this card if you don't have a good cpu. There is no discussion that more cards are good in the market we have right now. The card is 50% over MSRP right now at most retailers.
> Are you suggesting that you could use Raytracing at all in modern games on a low end card like the RX 6500 XT?
Yes?
Minecraft Raytracing and Quake Raytracing are rather low-specs and probably would run on an RX 6500 XT. You might have to drop down to 720p but it probably would run.
Are you kidding? These are fully path traced games and they are extremely hard to render. RDNA2 cards are worse in raytracing titles than Nvidia cards. For example 6900xt is half as performant in Minecraft RTX compared to 3090(18FPS vs 40FPS) according to LTT tests(testing was in 4k). And according to random tests I saw on Youtube. RX6600 outputs around 6fps with RTX enabled in 1080p, and it's basically 50% stronger. Not sure about 720p, but I would be surprised if it 5500XT gets to 10FPS at that resolution.
I don't think that's true - the GTX 1060, which was its' competitor, and matched it's performance pretty closely only had 4.375 TFLOPS of compute. Architecture matters
It looks like it is just a different card designed for a different purpose. High clock speed, more power efficient, and a lot smaller...and $200 (supposedly).
I have both those cards and your very wrong. The RX480 with 8gb still can’t match the RX6500 with 4gb. The RX480 was a great mid tier card, but it’s showing it’s age and the RX6500 is a good upgrade. It’s not drastically better, but enough to play most games 1080p and medium to high graphics. I was stuck at 1080p and low to medium with the Rx480.
Isn't it a power budget and heat dissipation thing? For example the Xbox Series X and Playstation 5 GPUs have pretty similar architectures, but the PS5 has fewer compute units at a higher frequency. Surely Microsoft would have loved to clock theirs as high as Sony.
PS5: 10.28 Teraflops, 36 Compute Units running at 2.23GHz (variable frequency)
XSX: 12.11 Teraflops, 52 Compute Units running at 1.825GHz (fixed)
> but the PS5 has fewer compute units at a higher frequency.
Yes, this seems to be what's happening here. Less power used on memory bandwidth, means more is available for other things. This card is not so much "bad at mining" and more like trying to design for a different niche that might be underserved by all the "good at mining" GPU's.
I love how Mark Cerny managed to spin Sony having to raise the clock frequency of the PS5 GPU to compete with a the XSX's TFlop specs as an advantage. GPUs tend to be memory-bandwidth bound - if you play around with raising the clock frequency by 10% from stock on most GPUs, you won't get 10% more fps, likewise if you drop the frequency by 10%, you won't lose 10% fps.
But with the power consumption scaling with the square of the frequency, the difference in power draw can be absolutely significant. That's the reason many people (me among them) chose to underclock their 5700XTs
My 1060 3GB can still run every game I throw at it (except Cyberpunk). MS Flight Sim runs great at high, for instance, especially after the performance update.
Looks like ~60 (more in rural areas, less in cities) for 1080p low on that card... which isn't fantastic but better than I expected. The limiting factor in cities seems to be the VRAM as the 1060 6 GB version can bench about the same flying through New York City bumped up to medium and shows using nearly 5 GB of VRAM while doing it.
For a flight sim, 45fps. Forza Horizon 5 runs at 60+ on high settings (despite telling me I need more memory, it doesn't crash). I mostly play CS, which runs at 200+ (yes I know it's an older game, Doom Eternal also runs at 60+ on high).
4GB ram is not enough for a good 1080p gaming card. You'd most likely only be able to play esport titles on the 6500 XT. I went from a 3GB 1060 to a 6GB 5600 XT and it makes the extra frame buffer makes a massive difference as new games are getting very demanding to run, even on 1080p.
I'm playing around more with deep learning as of late. Are the same technical specifications one would look for in a 'miner-friendly' GPU the same ones I'd look for in a GPU for Neural-net computations?
> there isn’t an 8GB version of the 6500 XT. The 6500 XT also uses a 64-bit memory interface
It only comes with 4GB of RAM on a narrow bus which makes it useless for Ethereum mining. It doesn't seem like it uses the same "LHR" technique as Nvidia.
This still doesn't explain why they cheaped out on the media engine side.
While I can see a justification for the lack of 4K H264 encode support (kind of pointless for a card at this level of performance), removing the AV1 decoder and H265/HEVC encoder is something I don't understand.
This makes the card completely useless for both game streaming and media creation for no discernible reason.
AMD consumer cards are unusable for GPGPU. Utter garbage. It’s not just mining. I would not have bought them if it weren’t for their open source driver. I hope Intel Arc fares better.
The higher-end cards are inherently very very usable for GPGPU, it's just the tooling around them that sucks. As for lower end cards, that's because they are low-end.
Yes, they have the Radeon Open Compute project (ROCM) but they seem to be intent on following the tensorflow from a few stable versions behind. Additionally, and likely this is not something they can do anything about, but if you are using a Linux VM instead of running it natively, ROCM does not work. I had attempted to do this to get some use of the AMD GPU that was in my Mac — no dice.
The holy grail would be a direct replacement backend that could be fed into TF, like CUDA.
If you are using a VM, you can never use your GPU unless you have a very expensive one or a spare GPU to passthrough.
ROCm has a direct replacement backend that can even take CUDA code (it's designed to be incredibly similar). It's called HIP. It's just that no one wants to support it. That is actually how TensorFlow on AMD works (mostly), and you can compile the latest stable release that way.
6600XT pass through works fine on QEMU/virtd. There is no need to have multiple GPUs in your machine to have a passable setup.
I can boot into Linux, and swap into Windows in 2 seconds with this setup. I have a dirty 20 line Bash script that deals with detaching the console, and passing the right things to the right place, but it all works.
ROCm on consumer cards does not work well. The tooling sucks. Massively. I don't understand why AMD doesn't have an extra team of 20 devs working just on the tooling.
Using DirectML with Windows Subsystem for Linux gives you better ML GPGPU support then AMDs native tooling.
>I can boot into Linux, and swap into Windows in 2 seconds with this setup. I have a dirty 20 line Bash script that deals with detaching the console, and passing the right things to the right place, but it all works.
That's a matter of opinion I suppose, but I don't personally find that passable.
>Using DirectML with Windows Subsystem for Linux gives you better ML GPGPU support then AMDs native tooling.
DirectML sucks even more than ROCm, IMO. Also, WSL sucks more than a normal VM.
What is a setup that would be passable then? I think a setup like the one that I have described [I believe] would be impossible with Hyper-V or ESXi (though, not that I have even attempted it with either).
I agree witht the latter part about Hyper-V or ESXi, for it to work one would need a virtualizable GPU or two GPUs, so that both OSes can have a graphical environment concurrently.
Yes and no. AMD uses different architectures for their consumer grade hardware and their HPC stuff as well (i.e. RDNA vs. CDNA).
CUDA support (or rather lack thereof, which isn't AMD's fault) plays a major role w.r.t. software support, but AMD's compute architecture isn't DL-focused either.
The MI250X compute part for example has a FP16 to FP32 ratio of 8:1 and a FP32 to FP64 ratio of 1:1; in other words it's an absolute beast at GPGPU compute.
The 6900XT on the other hand has a FP16 to FP32 ratio of just 2:1 and a FP32 to FP64 ratio of 1:16 (i.e. it's severely restricted at high precision workloads and OK at half-precision).
Comparing this to the specs of NVIDIA cards, they're still vastly superior on paper. The consumer versions of Ampere only get 1:1 (FP16:FP32) and 1:64(!!! FP32:FP64) respectively. But then again, NVIDIA cards feature dedicated "tensor cores", which have no equivalent on AMD consumer grade hardware.
The main selling point for NVIDIA, however, is software support and mindshare. They started to buy themselves into academia in the late 2000s by sponsoring labs and providing a vast ecosystem of software libraries for deep learning and GPGPU support in general. This not only helped kickstarting the deep learning revolution but also tied their hardware and brand name to GPGPU, which basically became synonymous with CUDA at that point.
Ah I see. But wouldn't that mean that hardware-wise, consumer grade AMD stuff is better than for GPGPU, than consumer grade NVIDIA stuff? With the exception of AI and software support, of course.
The previous generations of AMD hardware were super popular among crypto miners for that reason.
Most other software focused exclusively on CUDA, so GPGPU on AMD cards is not well supported despite the potential.
Blender is a popular example for this. After discontinuing cross-vendor OpenCL support, which effectively limited GPU acceleration to NVIDIA cards, they added support for AMD cards in the latest version. Even then, only the latest generation of AMD cards is supported for some reason.
Other renderers like OctaneRender still only support CUDA. The situation is just as bleak in video editing software, were major companies like Adobe only support NVIDA and Intel (at least on Windows) or have poor OpenCL support (which users then blame on AMD of course).
There is some hope that software support for GPGPU without CUDA (maybe using Vulkan Compute?) will improve later this year with Intel (re-)entering the discrete desktop GPU market.
Man, I'm glad I snuck a 1080ti purchase in a few years ago when BTC was crashing hard. Of course, being Nvidia means it'll suck if I have to transition to a Linux Desktop before the next cryptoshit market crash, but if it takes that long then they've probably managed to kill of high-end PC gaming along with the environment and I'll have bigger problems.
I'm afraid it isn't worth much, considering how many times a Linux Desktop user has told me "I haven't really had any problems with it" about something that ended up giving me no end of problems.
Looks like I bought it in June of 2018, and BTC was ~$6k, which was $7k down from its previous high in December of 2017. It would drop another $3k by January of 2019.
At least, that's what the first graph I found said.
And all crypto assets at the time. People wouldn’t be clamoring for GPUs if Ethereum didnt exist and also capture so much of the market. Its the only proof of work blockchain that people pay so much to use that it frequently exceeds the block subsidy.
so Bitcoin crashing, without Ethereum existing, wouldn't mean anything. thats how I separate the two without it seeming pedantic to me.
I started using Nvidia cards with Linux about two years ago and haven't had a single issue so far. It has always been Ubuntu, so that might be part of it, but I get the impression that Nvidia's beef with Linux isn't an issue anymore.
I have one machine using AMD and another using Nvidia. The only quirk with the Nvidia drivers, from my perspective, is having to install them separately under most distributions while AMD graphics just work. Even then, there is a lot less hassle to using Nvidia drivers under Linux than either vendor's drivers under Windows.
(I recognize that there have been issues with Wayland support in the past, it creates issues for some developers, and support for modern Nvidia cards on most other operating systems is non-existent. There are likely other issues, but it seems like most of the issues would affect a small number of people.)
For general desktop use, it's mostly fine. However, for gaming I noticed way better performance and no more graphical glitches after switching from a GTX 1060 to an RX 570 (a worse card on Windows, especially OpenGL). Gnome seems to handle Nvidia better than KDE too.
World really needs a ban on PoW cryptocurrencies. I guess on one side lot of people is not caring about it as long as they make some money of it, and in the other side maybe other people in power, is not caring as that would make a perfect excuse to ban any kind of cryptocurrency. I guess too we all will care as soon as the devastation caused by the climate change becomes a massive life threatening issue.
You are moving the goalposts. If Bitcoin miners pay for renewable energy, they are no longer harming the environment.
Who are you to dictate what someone does with energy they pay for? Just because you do not like what they do, doesn't mean you can ban it if it's not harming you.
There is no evidence that Bitcoin's usage of renewables is preventing others from paying for renewables too.
No, that's not moving goalposts. If Bitcoin miners waste renewable energy, the users of energy in general move to renewable later.
If everybody is using renewable energy exclusivity for everything, is still wastage. PoW only work by adjusting it's consumption upwards to create scarcity, e.g. one block per 10 minutes. At that point we will be useless solar panels, covering land uselessly, etc, etc. Bitcoin is arbitrary scalable wastage to create arbitrary, artificial scarcity.
The point is: Energy wasted is energy wasted. Bitcoin will eat everything till the point where some external factor prevents further expansion, e.g. how much are we willing to hurt the environment and ourselfs in exchange of the block reward. The person with the highest threshold wins.
It's not really not complicated. This is literally the last scene of the hitchiker, where they plan to kill all trees, so they can use the leaves for currency. The success of bitcoin makes be loose all hope in humanity.
They explained the difference for you in the second paragraph: unless any other system, proof-of-waste systems are uniquely bad because the amount of power used isn't linked to any sort of useful activity — it's just a function of how much competing hash capacity there is in the network. Even something like heating your house to 78℉ is a lower category of waste because you won't increase that to 120℉ or start heating the neighborhood park just because your neighbors matched you at 78℉.
This is also in keeping with tons of existing regulations: you can't buy a monitor which uses 2,000W because various governments have caps for how inefficient you're allowed to be. That's far from perfect — or sufficient to deal with the climate change crisis — but it's not like it's some sort of unprecedented imposition on personal freedoms to cap inefficiency.
so you made a judgement on what an activity is useful. There's no such objective judgement possible, since usefulness is a subjective matter.
Therefore, pricing carbon correctly (and thus pricing electricity correctly) is the way to provide all parties the full information on cost, and each would decide how best to use their own money to buy energy.
> so you made a judgement on what an activity is useful.
No, I understand how the system works. Bitcoin uses power as a function of the aggregate network capacity — it must to avoid a 51% attack — and that means that the power usage not coupled to the number of transactions being processed: even if Bank of America wasn't already many orders of magnitude cheaper, they wouldn't buy a mainframe they didn't have work for just to stay tied with Wells Fargo. It's the computational equivalent of Luftansa flying all of those thousands of empty flights.
> Therefore, pricing carbon correctly (and thus pricing electricity correctly) is the way to provide all parties the full information on cost, and each would decide how best to use their own money to buy energy.
This would help considerably but it's not enough: we need to radically pivot the global energy system and that will involve things like shifting where major power consumption occurs (e.g. ideally things like blast furnaces or smelters would be located near sources of hydro, geothermal, nuclear, etc. power) and getting everyone to convert to things like EVs en masse.
Unfortunately those decisions take a longer timeframe than shipping a container full of cryptocurrency mining hardware to wherever the current cheapest source of power is. That adds a lot of noise to the pricing signals we need to drive business decisions, especially for major capital expenditures which require years of planning. It doesn't help if, say, Iceland adds 10GW of clean geothermal power but the price stays the same or even goes up because a bunch of miners showed up the week after it went online.
>"Waste". The parent's point is who are you to say?
You didn't read or you didn't understand what I said. It's waste because it's designed to create a scarcity. Not by the judgement of anyone of the usefullness. The power bitcoin uses is strictly a function of how much power is available, not what service is provides. Double the number of transaction rate: No change in power usage. Half the price of electricity: Double the power usage. This strange lack of connection between utility and resource usage makes it wastage. It might be hard to grasp, because there is nothing unlike it.
> Maybe we should closely audit your energy usage and ban those activities we find objectionable.
As stated above the energy consumption of bitcoin is not bad because the a decentralised ledger is objectionable, but because it's a controller (in the sense of control theory) tuned to make energy a scare resource. This is a scheme equivalent to poison every aquifer in the world to profit from the scarcity of water. It's doesn't matter if scarcity of water is useful for some apparatus, or if the task the apparatus performs is objectionable. If you want to monitor my non bitcoin energy-usage demonstrates you didn't understand is difference between utilisation of water for the sake of using it and for the sake of wasting it. Usage can be made more efficient, and new resources can cure a lack of water. In the case of wastage the wastage can by definion not become more efficient and new resources will not make it last any longer, but simply call for an increase the rate of poisoning.
Even if bitcoin had any use case besides randomware, which it doesn't, even if it did solve every problem the financial world had, it would see be wasteful. While most other technology becomes more efficient with further use cases, PoW becomes more wastefull, for reasons state above. The only way to make it more efficient is to make energy less aviable to humanity, as such are the perverse incentives of this scheme.
No, I'm recognizing that the economy is connected. Many of the best renewables — especially those which are available 24x7 — are things like geothermal or hydropower which have finite capacity, and non-trivial ramp-up time.
Proof-of-waste systems like Bitcoin are uniquely inefficient due to the way the network adjusts to increase the difficulty — an aluminum smelter uses a ton of power but they don't build new ones in excess of market demand for aluminum — and they're easy to ship around the world. That means that demand always goes up and any time some place has cheaper power for a non-trivial period of time someone will probably ship a bunch of mining hardware there.
This is a problem because right now we need the entire global economy to electrify as soon as possible. That means things like industrial processes need to be relocated to cheap sources of clean electricity because you can't just plunk a coal power plant next door, etc. and there are also challenges with peak power demand requiring upgrades to electrical grids or other equipment. Having so much power dedicated to something which has never found a useful purpose is a source of friction we don't have time for.
>> Just because you do not like what they do, doesn't mean you can ban it if it's not harming you.
> Believe it or not, this is not how society works.
and yet, we are now unbanning a lot of activities that some people dislike - such as smoking pot. the OP's point is that not everyone is in consensus about the energy use of PoW being a waste.
any infrastructure dedicated to the mining of crypto is opportunity cost. Even if you use renewables, that energy could instead replace traditional energy sources which could be turned off.
This is an example of Jevon's paradox, namely that increases in efficiency or quality are offset by increases in demand. It's one of the reasons why we have globally barely made a dent when it comes to the increase of renewables as a total share of energy consumed despite the fact that the capacity has increased pretty much exponentially.
This is made even worse by the fact that green infrastructure itself necessarily often consumes non-green sources, in particular during construction.
> that energy could instead replace traditional energy sources which could be turned off.
This is not true. Often, renewables aren't available to some locations but abundantly available in others. Other times, some people buy electricity from renewable providers and others don't.
With 50-75% of Bitcoin mining being renewable, one thing is clear - it's helped the adoption of renewables and made them cheaper.
It has helped the adoption of renewables in the same sense how polluting the local river has helped the anti-pollution industry. This is broken window logic. The reason bitcoin uses renwables in the first place is because much of it happens in China where misguided state projects and subsidies have created excess capacity. It makes for a pretty decent joke though, the libertarian currency de jure runs on the misguided state planning of the Communist party of China.
Any excess capacity spent on crypto currency is waste. It could go to scientific computing projects, it could simply not be used. The chips for the equipment (we are in a global chip shortage) could go to something that actually contributes to the real economy.
It may not necessarily be the same as "broken window" logic. Consider a world where no one had ever broken a window. There wouldn't be any window repair people. The first time a window breaks, we have a new opportunity to fill a need. Necessity is the mother of invention. The exponential demand for energy creates more pressure to develop more efficient renewable energy generation methods that probably wouldn't be incentivized by a constant amount of energy use. It's a risky bet, but it might be the only way to spur humanity into action.
We just need to increase carbon taxes at the same rate that renewables become affordable.
it's a complete non-argument you can use it to justify literally anything. Dropping atomic bombs on Japan, hey construction boom! Covid-19, pharma industry innovation! Do you turn the heater on and leave the fridge open to accelerate humanity's fight against climate change? No.
We have about a few thousand pressing issues that spur us on to push for renewables as it is, we don't need to waste energy on purpose.
> We have about a few thousand pressing issues that spur us on to push for renewables as it is, we don't need to waste energy on purpose.
Crypto mining is one of those issues, we don't have a choice. Your argument reduces to "if we had no issues pushing us, we would have enough issues pushing us" which is a contradiction.
"...helped the adoption of renewables and made them cheaper." How so?
Miners are chasing profit margin so they'll use whatever's cheapest. If that's renewable where some of them are, so be it. And if everyone else start reducing use of coal and that becomes cheap, miners who exploits that to derive higher profit margin will grow faster I imagine.
I would attribute growth and economies of scale that renewables have achieved to renewable proponents, industries, government subsidies and loan prorams. Crypto miners are energy consumers and some of them use renewable energy. But how do we go from there to miners being some kind of force for good in this context?
> This is an example of Jevon's paradox, namely that increases in efficiency or quality are offset by increases in demand.
That is kinda the main point of a carbon tax... it is to make carbon usage LESS efficient, since you have to pay an extra tax in addition to the cost of the energy itself. Ideally, the money gained from the carbon tax is then redistributed to everyone equally, and that money can then be used for whatever you want.
> Ideally, the money gained from the carbon tax is then redistributed to everyone equally
i dont think that's a good idea - the redistribution does not decrease carbon.
I think the correct carbon tax is the amount of money it takes to fully remove the carbon emitted (at current tech levels). This pot of money is then used to actually remove carbon.
Yes. That's why you need big players - like the US, EU - to use leverage to force the rest of the world to follow suit. But the change has to start somewhere.
That solves nothing, it only lines politician’s pockets. The cost of electricity is the deterrent currently, which is why you don’t see everyone mining. If you want to make electricity cost more, it’ll “solve” more than crypto mining, it’ll just make mostly poor people have a lower standard of life.
The best carbon fee proposals refund the money to taxpayers, equal amount per person. If you emit less CO2 than average, as most poor people do, then you come out ahead.
It doesn’t make much sense, how would money be guaranteed to come back after making its way to through the government and back? The cost of electricity already functions more efficiently; don’t take money out of peoples pockets so you don’t need to give it back.
A pigovian tax by insurance would work better, I definitely think environmental protection is needed but politicians will just use the money they get in nefarious ways.
How can you guarantee that people will get their tax refunds? Their social security payments? The carbon rebate is easier than that, since everybody gets the same amount. Don't even pass the money through the general fund, where it's accessible to politicians.
Collection can be simple too. Charge at major sources like coal mines and oil wells, and the fee gets rolled into the prices of everything downstream.
All a carbon tax does is punish consumers, while making crypto miners make less profit. The amount of tax it would take to make crypto mining less profitable basically is an imposition of austerity.
That's not how this works. Even if you pay your supplier to buy only renewable energy on the open market, if your house is supplied by a coal power plant then your mining rig is powered by coal, regardless who you're paying for your electricity. Yes it feels good to pay for renewable electricity but ultimately that's not how it works.
And crypto needs to be banned because it's all a gigantic scam that has a gigantic environmental cost attached to it.
And my emissions are offset for the fee that I pay.
Carbon credits, more planted trees, whatever. My net impact is likely lower than anyone that doesn't pay for renewables, even with mining.
> And crypto needs to be banned because it's all a gigantic scam that has a gigantic environmental cost attached to it
For PoW, it's the purest form of currency - the value of energy itself.
For other currencies, it's no different than precious metals.
Trust HN to call decentralized technologies enabling pseudonymity/anonymity scams on one hand while advocating for decentralized technologies enabling pseudonymity/anonymity on the other hand.
Also, from my perspective, you should stop eating meat. Your eating of meat provides you nothing but selfish pleasure while causing a ton of carbon emissions - and actual tangible suffering on scales never before seen on this planet. The magnitude of suffering caused is tens billions of times worse than crypto. Let's ban meat first.
Those do something useful and there's a cap on how much people use: if I put up lights and then my neighbors put up lights, I don't put up twice as many lights so I can out-compete them for carolers and I certainly don't buy another house to put up even more lights. Proof-of-waste effectively requires that of all participants.
People (and cities) absolutely try to out-compete each other with christmas lights. Even countries with negligible Christian populations are taking part nowadays. Don't forget all the other wasteful elements of christmas. And all the christmas-related scams and crime. And the yearly spikes in domestic violence. The growth of the christmas price index seems to go to the moon. How is celebrating some guy's birthday 2000 years later useful and how does it justify all of this? And does it at least really have a lower environmental impact than crypto?
There's some competition, yes, but it's bounded — even the most extravagant Christmas display is nowhere near the maximum amount of lights which could be placed in the space, not to mention only operating for a few weeks a year.
The design of Bitcoin ensures that does _not_ happen there, and it's currently using a massive amount of power to process a tiny fraction of the total economy, especially after you account for the transactions which are not linked to real economic value (money laundering, wash trading, etc.). The design ensures that if someone did find a useful application which saw popular adoption, usage would need to be many orders of magnitude higher than it is now.
That is how it works though. If my money is only allowed to be used to generate renewable energy, then it doesn't matter how the power that comes to my house was generated. Energy is energy. If everyone does what I do, then no one will be left to fund nonrenewable energy. But the only way that would happen is if the cost of nonrenewable energy went up proportional to the damage it causes via a carbon tax.
And banning crypto mining is impossible. It's arbitrary math. You can't ban a computer from doing arbitrary math. At best you could discourage businesses taking crypto as payment, but even then people can operate 100% in crypto and avoid any regulations.
Buying renewable energy isn't the same as purchasing carbon credits. Since energy is fungible there's no way to say your demand for renewable energy actually mattered if others are agnostic in their demand for renewable versus non renewable energy. When we use more energy it's not renewables that are picking up the slack, the fossil fuel plants burn more.
Imposing global technological bans so willy-nilly seems like a nice idea only on the surface. Things like machine drying your clothes at home in the US (uncommon in Europe so seems frivolous) uses as much if not more energy. Others might consider gaming or self-hosted servers or who knows what else equally inefficient. Do you really want to put things so easily on the potential chopping block even if this one thing you hear a lot about might be a waste?
Now if PoW was using 10% of the world power maybe it'd worth it but I just don't see how you can justify such a ban but only apply it to this one thing.
Machine washing is a significant labor and time saver.
People are usually opposed to line drying clothes because of aesthetics and feeling like it makes their neighborhood look “poor”. Luckily quite a few states have passed laws making it illegal for landlords or HOAs to ban putting up clotheslines.
Significant? Who doesn't have 20 minutes to soak and wash their clothes in a bucket? If the goal is to reduce energy usage one should start at the top with large appliances.
The typical drying machine pulls about 2-6 KW while running. A cryptomining rig pulls about ~1.2 KW while mining. A drying machine runs for an hour a week. Let's say 2 hours a week. That's 12kWh per week at the high end for drying. A cryptomining rig running 24/7 ends up using 201 KWh for the same time period. That's 16x more taking the most pessimistic clothes drying usage, so I'm not sure how you're concluding "uses as much if not more energy".
Most people living in the suburbs in the US have a drying machine while not everyone has a cryptomining rig running 24/7, so we could _easily_ be in a situation where drying machines in a region use more power than cryptomining rigs. And a lot of things are like that: Netflix, furnaces, AC, freezers.
You forgot to account for the cost of the damage a dryer does to clothing. And how much conditioned air it pumps outside that needs to get sucked back in through my home’s cracks and re-conditioned.
OP is probably talking cumulative energy. The vast majority of households probably have dryers in the US. The vast majority of households also DO NOT have crypto mining rigs.
It's currently close. Bitcoin alone uses 91 TWh globally [1]. US dryers consume ~60 TWh [2]. However, one of those is seeing exponential energy usage while the others is very mature and stable.
I'm also shocked that people do laundry so often. Not sure why people are doing more than one or two loads per week and that's a problem the government could look at tackling (not so much with PoW coins).
A household dryer is very unlikely to get anywhere near 6 kW even at peak. In Europe it's common to have 16 A at 220-240 V which gives you a maximum of roughly 3.5 kW.
I just looked up how much energy my last dryer load used. It came in just shy of 6kWh. The unit is a standard electric dryer, around 10yr old. Just wanted to provide some additional data.
We do about 4 loads per week here. I hope that the newer heat pump based dryers are reliable!
Biggest problem about PoW coins, and that's something that everybody working on it knows, or should know, is that the need of power is steadily bigger each year. Where would be the limit of that? When that requires the energy of a big country? (We are quite away from that as far as I know [1] ) A heavily industrialized continent? More than the whole world?
It's absolutely crazy. I know that there's a big lobby behind mining, I absolutely respect what they have being doing protecting Bitcoin i.e. and making the project survive... even if it somehow completely lost it's original purpose. But FFS guys what's the purpose of having cryptocurrencies if the cost of that is aiming to destroy the world, or much more probably, to win a last minute complete cryptocurrency ban from all governments with the handy climate change as an excuse?
>Biggest problem about PoW coins, and that's something that everybody working on it knows, or should know, is that the need of power is steadily bigger each year.
They don't really require more power, they are just getting more valuable so it's been profitable for more people to increase their mining investments. The limit is that they are not infinitely valuable, and if anything mining rewards are decreasing.
>Biggest problem about PoW coins, and that's something that everybody working on it knows, or should know, is that the need of power is steadily bigger each year
PoW is wasteful by design. The more hash power you throw at the network, the more difficult the PoW becomes. Any efficiency gains necessarily have to go towards more hashing, instead of maintaining the same level of work at a lower power draw.
I sometimes wonder if we'll discover a Kardashev Type II civilization, harnessing the power of their sun via a Dyson sphere. And then, upon first-contact, realize they are using all that power to mine Bitcoin.
Indeed. What then, do we get them to move to SHA256 (or whatever hash Bitcoin uses for its PoW in 100 years time), or do we adopt the alien hashing algos?
It doesn't make sense to compare 1 person gaming or clothes drying to a specific mining machine. The consideration was about global bans and total outputs from those activities.
I have no love for bitcoin, but I always hate this comparison. It’s useless. Sweden uses much more energy than it uses power (as pretty much every country) and it also exports power usage by buying finished products. Why is the power usage not compared to something useful? The power or rather energy usage of the global banking network for example?
When people compare energy usage like that, they are making an implicit argument that the energy usage of bitcoin is a "waste".
This is what i dislike about the argument that PoW should be banned - that it wastes energy that could otherwise have been used for something else. I argue that it is not a waste, because the person who is paying for the energy does not consider it a waste. No one else can make this judgement.
The only weakness of this argument is that some energy sources have externalized the cost of emissions. I would very much like to see an emissions tax to account for it (which makes sense to do, regardless of one's position on PoW).
Agreed with the "waste" argument. I won't defend the huge power consumption but the energy spent is strictly useful in a utilitarian sense - every joule spent makes it that much harder to perform a 51% attack. Ideally, the energy used to secure the network is widely/evenly distributed among nodes for maximum security benefit (though not usually how it plays out).
However, to say it's all a waste is just an emotional argument because you don't like it. That's fine but it's opinion, not objective fact.
> When you put it all together, that’s a projection of 135.12 TWh in the year, or about as much power as is used annually by the country of Sweden
I don't doubt that bitcoin miners use a lot of energy, but I hesitate to trust a source that doesn't know the difference between power and energy. Also, they're lying about the energy consumption of Sweden, which actually uses 645.7 TW/yr as of 2012 [0]. They number they quote is actually the electricity consumption of Sweden [1], which is like 5x lower.
Lets say Coinbase and Paypal (both BTC processors) stop supporting your website. Do you think that you'll get as many BTC transactions if this happens?
I don’t think you understand how cryptocurrencies work. There ain’t much Coinbase and PayPal can do to stop someone sending coins to my address. I don’t have a website, nor do I care what Coinbase or PayPal do or say.
I don't think you understand the practical business realities of the internet.
No typical business is going to work with BTC addresses on the raw. They're going to (and currently) using simple APIs like Coinbase API (https://developers.coinbase.com/) to "implement BTC receipts".
As far as US business is concerned, the typical payment processor for bitcoin is Coinbase. They handle the bitcoint transfer, transaction history, and conversion into USD all at the same time.
There are many business that run their own payment infrastructure, BTCPay is the most popular. It is better for the network if people DONT use centralized providers. We all have the power to use decentralized infrastructure.
Bitcoin works the way it is intended, that is peer to peer payments. Everything else is just ecosystem, which can be replaced. Coinbase is important as an on-ramp for users for now, yes, but their importance to the network is peripheral.
Bitcoin doesn’t care about businesses. It doesn’t care about peoples wishes. It doesn’t care that everyone else is using Gmail. It provides a peer to peer, censorship resistant append only ledger for transactions secured by cryptographic signatures and proof of work. That is all. And that is plenty.
I agree. I think you two are talking past each other to a degree - Bitcoin the protocol is decentralized, Bitcoin the ecosystem in 2022 is moderately centralized in various ways.
Without an on-ramp and off-ramp for converting into dollars/euros, Bitcoin is entirely useless, even for the hardcore idealists like yourself. And those can quite easily be made illegal.
Maybe in eth land, if you don’t run a fullnode. And that speaks to moxies recent post about centralization.
I’m talking about Bitcoin, and you better believe I run a fullnode. No provider required. I recommend you run a Bitcoin full Node too, and transact without any 3rd parties!
I think the issue is your concern does not really answer the question I posed. We can follow the use 'use' argument as soon as we deal with the 'who has the authority to tell people what to do with energy/money/equipment'. Is it me? Is it you? Is it society? Is it government? As soon as we get through that conundrum, we can discuss how crypto is basically evil and used for nefarious purposes only and cars are more like knives ( I disagree, but we can get to it once we know who can actually make the call on the behalf of humanity as a whole that a given technology is banned ).
I don't understand why anyone thinks this argument is some kind of checkmate. This is the equivalent of screaming you're a sovereign citizen when the police ask for your name.
Governments have this general monopoly on the use of force, which is why they tell you you can't do things like sell heroin. If you disobey them, they will send men with firearms and make you stop.
Governments can obtain their consent in a bunch of different models and lots of people argue about which ones are better.
A "world wide ban on crypto" probably is not going to happen but more and more countries are banning it. I suspect many more will continue.
You still did not give me an answer by the way and I have to question as to whether you are arguing in good faith.
Still, I will address your comment in good faith.
It is not a checkmate. I am simply just trying to see what world model you have in your head. We clearly disagree, but it does not mean we can't openly discuss those disagreements. It is mostly the beauty of HN. If you present a logical argument, I might simply say, it makes sense. Right now I am confused.
Are you officially arguing government is the one to say as to whether technology is banned? If so, which government speaks for humanity as a whole?
<<This is the equivalent of screaming you're a sovereign citizen when the police ask for your name.
Is it? Can you elaborate a little bit? The analogy seems a little off.
<<Governments have this general monopoly on the use of force, which is why they tell you you can't do things like sell heroin.
Yes. Government monopoly on power has been a well established theory. We are not breaking new ground here. Is crypto heroin in this scenario?
<<If you disobey them, they will send men with firearms and make you stop.
Yes. See above. How does that help your argument?
<<Governments can obtain their consent in a bunch of different models and lots of people argue about which ones are better.
Yes. See above. How does that help your argument?
<<A "world wide ban on crypto" probably is not going to happen but more and more countries are banning it. I suspect many more will continue.
Maybe yes. Maybe not. It is not, however, an argument for or against crypto banning, whether it should be banned or not. It is not an argument at all. At best, it is just a speculation.
I'm going to interpret your current question as "Who has the authority to ban crypto?"
The answer is the government.
Governments are restricted to their borders but they do like to cooperate, and so EU may pass a general crypto ban and the EU may even ask the nations it trades with or that want to join to do similar.
But there is no real question that a government may ban something in it's borders.
"Ah, but I can get away with crime sometimes! Thus the government ban doesn't matter!" is the usual response. And sure, some small crime goes unpunished, but in general people don't want to overtly engage in criminal behavior. No business would touch crypto, which harms it's value and use.
Should we ban crypto?
Well, we can explore why countries have. The answer is a mixture of "it makes tax evasion stupidly easy" and "it uses too much power". With parts of the world paying INSANE prices for fuel right now, a lot more governments will latch onto that second line of reasoning. In general, governments detest tax evasion and things which aide it.
We do in fact generally like tax revenue. We like roads and schools and such, and most people are opposed to tax cheats. Nobody particularly cares for high electricity prices or polluters either.
"Well, what about those crypto operations which pay their taxes and are funded by renewables?" Sure, perhaps those are allowed to exist, but that comes with two problems. 1. It dramatically shrinks the mining pool, making sybil attacks easier. 2. It reduces processors which means less businesses will accept crypto.
Crypto already has a strong association with open air drug markets, and for good reason. As legitimate usage declines, the desire to ban a tool used primarily for crime will increase as it should.
Crypto is already being banned and it probably should be banned. If not banned then regulated, but regulation would harm it significantly.
<< Well, we can explore why countries have. The answer is a mixture of "it makes tax evasion stupidly easy" and "it uses too much power". With parts of the world paying INSANE prices for fuel right now, a lot more governments will latch onto that second line of reasoning. In general, governments detest tax evasion and things which aide it.
I take issue with the statement "it makes tax evasion stupidly easy". The moment you touch any US-based exchange, your information is up for grabs by a variety of agencies from IRS to various LEOs. It may make doing the deed easy, but it does not really change much. It has been years now since blockchain made it into mainstream. There are companies specializing in selling blockchain insights. Unless we are not talking US. But US has an entire cottage industry devoted to making sure that the wealthy pay as little as possible. I am not sure banning crypto would help here. Killing lawyers and accountants could ( I jest.. it is just a reference to Henry VI ), but that is about it.
<<"it uses too much power"
I find this argument amusing. I am paying for it. If you don't like my use, either raise the price or, well, leave me alone. But therein lies a problem, energy prices are not an equilibrium point and people do not want to give up THEIR toys, but an appropriate scapegoat will let them keep their toys and feel righteous indignation. It is a win/win.
<<The answer is the government.
And this is where it gets fun as correctly noticed. Even without considering the distributed nature of the technology itself ( because there are way to circumvent that as internet splintering has shown ), government these days may, if not will, disagree for a variety of reasons including but not limited to "they are for A, therefore I am for not A", making it a protracted game of whack-a-mole ala piracy, war on drugs and multiple other endeavours governments proved to be so good at. To sum up, you may reasonably argue they have the authority, but do you reasonably think, they will actually pull it off?
I find this argument odd. Is the electricity used by the business somehow different than the one used by grandma? Are these currents used by the business somehow more special? No. If anything, based on my dad's experience as a shop owner, businesses pay much higher rates than individuals. They are not subsidizing grandma. They are subsidizing bureaucrat lives. That is all the there is to it.
Still, let us say that I accept your argument, some miners actually bought various electricity producing entities. Now what? Are you going to argue they can't use the electricity they generate or are you going to argue that we can then force them out of business some other way. Either undermines some tenets of US society, but I am curious now.
It may make doing the deed easy, but it does not really change much
I'd argue the opposite, bitcoin is terrible for tax evasion and only fools think they're getting away with it. Besides every legit exchange doing KYC, law enforcement buys off the shelf chain analysis software to track people on their radar.
> "Who has the authority to ban crypto?"
> The answer is the government.
The gov't is supposed to represent the people's will. So are you saying that it is the will of the people to ban crypto? Aka, is crypto something that is causing harm to society, like heroin addiction etc?
Let's go back to the prohibition era - why did the gov't decide to ban alcohol? And why did they re-instate it back afterwards?
I argue that gov't is not an "authority" - they merely execute the ban. The authority lies with the average views of society. Bans on slavery, for example, is only possible because most people slowly came around to disliking it (mostly people who don't benefit from it - while those who do vehemently fought against the banning).
Was it the "will of the people" to ban all drugs? Because that's not how I remember it. Specifically one good example is how MDMA got banned, tons of psychiatrists and psychologists went to court to speak out against the ban and the judge agreed with them. Did the DEA give a shit? Absolutely not. Banned.
Not to mention, the US forced the anti drug ideology on the rest of the planet through economic coercion
who is the authority? We, at least in democratic countries if enough people decide we should ban crypto we just draft the laws and end it, easy. We did this with many other things, don't know why this would be different
It is not a bad argument, but 'we' seem to be disagreeing a lot lately. I am not assuming you are from US, but I am going to use US as an example. The polarization is at a point where just about every major issues is split 50/50. And this does not even begin to touch as to whether US political class actually represents the needs/demands/whatever you want to call it of its electorate.
So yeah, you are not wrong about the general framework, but do you think drafting such a law right now would be..hmm.. palatable to the general populace. After all, consent of the governed is a thing.
And as for crypto specifically.. well.. the time to squash it has long passed. There is people with real money in it now. The best you can hope for is regulation.
> I consider trucks in US for non-contractor use to be frivolous, but I don't go around telling people those need to be banned.
What does you not liking trucks have to do with the issue of putting limitations on mining crypto?
Just because you don't care about this random issue doesn't invalidate serious concerns other people have.
Also, this is a pretty bad faith argument to say 'who can make these decisions'. We make rules in society all the time. If enough people can be convinced for a particular regulation and the laws of the society allow it, then it'll be implemented.
At the end of the day, trucks are there to illustrate a point that people may not agree what is considered 'good','efficient' or whatever qualifier you want to add. Trucks - those humongous behemoths intended for largely for contractors, but bought people with more money than sense for reasons that are not 'good','efficient' or whatever other qualifier you would want to add to crypto - are largely inefficient for a family unit. Some could easily argue that those should be banned for personal use as their gas-guzzling abilities are a MUCH greater threat to climate change than the entirety of crypto mining operation.
But do I suggest people are mandated to only get one Prius every 5 years? No. Do you know why, because the concerns of the people may be real, but we have not, as a society, yet decided that it is ok for those concerns to override individual choices.
You are right about society. You are wrong about the argument.
1000W+ power supplies are pretty common on the high end of gaming PC's, I've got a 1200W psu in my machine, as my old 750W PSU was shutting off when I upgraded my GPU. I don't know how many watts it's drawing during a resource intensive game, but it's gotta be a lot more than 400.
I think I'll look into getting a wall plug power meter.
RTX 3060 + 5950x CPU under load is like 400 to 450 watts, based on my power meter. That includes power to monitor and other peripherals as well.
I agree with OP though, we're within an order of magnitude of other legit uses, so let people be free. Increase electricity costs or taxes if you want.
A modern dryer uses less than 2 kWh for an 8kg run [0]. That means for those 38 kWh you need to run it daily for almost 3 weeks. Normal usage (let's say twice per week) would lead to ~3 kWh energy consumption.
[0] Siemens WT47R400: 176 kWh / year for 160 cycles by energy label, 1.4 kWh for a full run.
You severely underestimate the power use of proof-of-work blockchains. It's on the scale of nations.
But the main difference is that whatever you compare with those blockchains, it will never have the critical flaw: the automatically increasing difficulty. Your dryer doesn't periodically flood the drum to make it harder because more people are interested in tumble drying.
However, market forces guarantee that the difficulty will go up to the point where it's infeasible to add more hashrate, e.g. because it becomes uneconomical or there's simply no more hardware available. PoW workchains must work at maximum inefficiency. (This bears out in the data, too, I think. BTC had its highest hashrate at the beginning of this month, when energy price were at an all-time high and BTC was at the lowest over 12 months)
There are plenty of other wasteful activities that should be limited as well.
Classic whataboutism. First of all your entire comparison makes no sense, you have to compare alternatives for use. What's the alternative to a dryer. Using a line and hanging your clothes. Fair enough, what about in the winter in New York? A ventless dryer? Doesn't use substantially less energy. And so forth.
For every conceivable use case there exists alternatives that use orders of magnitude less energy. NFTs, sending money, federated computing, you name it, something already exists.
The only reason blockchain is a thing is because cryptocurrencies are booming in terms of how much they're sold for. The end.
> Fair enough, what about in the winter in New York?
I dry my clothes on a line in the winter in New York. It's below zero outside, but like 80 degrees and 0% humidity inside my apartment. Clothes dry fast. (And the space was already being heated, because I live in it, and it's against the law for the landlord to allow the temperature to go below 55 degrees, and they are paranoid, so it's more like 80-90 all winter.)
Sometimes I feel like I have a fever and am getting sick, then I realize it's probably just in the 90s and I should open a window. Not great. This is with all my radiators turned off!
The heat is on a timer so it's not like this all the time, but I have to wonder what a world where thermostats have been invented would be like.
>Fair enough, what about in the winter in New York?
Large swathes of Europe have winters as bad or even worse than New York's. I've grown up in one of those countries and had never seen a home clothes dryer before visiting the US. I've also almost never seen them in London or Berlin where I lived after that and those have at most marginally warmer winters than New York.
Not all Europeans use a drying rack indoors. In the Netherlands only the poor do this. It takes a long time and doesn't even save any energy as the water from your clothes makes your entire house humid and feel cold in the winter.
Governments could apply a carbon and electronic waste offset tax to declared cryptocurrency income. The people making money from cryptocurrencies should pay enough to fund projects that will fully offset their environmental costs.
Of course, those taxes will probably just disappear into general tax revenues and wind up paying for other things, but it would encourage crypto miners to pursue green energy sources and, perhaps, find better uses for obsolete graphics cards.
> The people making money from cryptocurrencies should pay enough to fund projects that will fully offset their environmental costs.
why should crypto be specifically targeted for an energy use tax (ala, a carbon tax but with extra steps), instead of a general tax on energy use (aka, just implement the aforementioned carbon tax)?
Would you support Visa and Amex if their next generation networks required you to burn x minutes of natural gas or x pounds of coal to complete a transaction? Why is this an acceptable architecture for Blockchain? It is a race to the bottom.
This comes up every single time banning cryptocurrency for wasting electricity comes up. Using a clothes drying machine and just about anything else that you can think of that uses electricity is fundamentally different than PoW cryptocurrency.
If I want to use a machine to dry my clothes, I pay for that electricity. This puts a limit on the amount of electricity I can consume to do it.
Cryptocurrency is different. The more electricity you burn, the more money you make. I know this first hand - the computer I'm typing this on is running gminer to mine ETH on flexpool.io with my RTX 3070 card. I've mined $2,800 of ETH in the past 10 months.
Something that most people don't understand is that there is a finite amount of ETH that can be mined per day. Hashrate determines the size of the slice of that pie you get. Therefore the incentive is to consume infinite electricity! Of course miners can't do that because mining gear (i.e. GPUs, ASICs, FPGAs, whatever) is hard to purchase and electricity isn't infinite.
But that doesn't stop miners from trying. I know people that do anything they can to acquire GPUs - they have bots combing the Internet, they buy whole systems that have GPUs, rip out the GPU and sell the system, they've signed up for every GPU manufacturer's wait list, they stalk Microcenter, whatever they can do. They seek out areas with cheap or even subsidized electricity and latch onto it like a parasite. It is nothing like anything else you spoke of.
> The more electricity you burn, the more money you make.
there will be a new equilibrium - the more electricity you use, the less of it is available, and thus the price would rise to match the demand. This means there's a limit to how much you can earn (unfortunately you wont know this limit until you reach it, because the full demand for crypto is unknown atm).
But high prices of electricity would spur developments in energy production.
In any case, energy use is energy use, regardless of the purpose. As long as the person using it is paying for it, then it is a valid use.
There is literal legislation about energy efficiency of consumer goods (such as dryers), our houses have to be built to be environmentally friendly with good insulation to stop wasted energy, governments are banning plastic packaging. But crypto should be except? No, crypto should be legislated so that it’s energy usage is limited. To think otherwise is an enormous step backwards.
I cannot understand how this is not a bigger issue being talked about in the media. Right now in the UK we are going through an energy crisis where people can’t afford to heat their own homes (I’m not blaming crypto for this). It’s absolutely maddening that in a world where climate change and energy scarcity is such an issue this would be a controversial topic.
My dishwasher is rated based on specific cycles, but I'm legally allowed to use it as frivolously as I want. I can run an A+++ rated dryer with a single sock in and send the photo to the police.
We don't limit energy use but inefficiencies. The equivalent in crypto would be to rate hardware by hashes per joule.
Crypto things are different because the incentives are different. I'm incentivised to use my appliances efficiently as the energy cost is just a cost.
On the other hand with PoW, I'm incentivised to mine efficiently but as much as possible because I can buy a thing that turns electricity into cash.
We are looking down the barrel of a gun. The gun is cryptomining's thirst for electricity destroying all our gains to diminish climate change. Your arguments may be right, but you are on the wrong side of our survival as a civilization.
The point here is that existing legislation does not cover this and trying to apply it would fail, you would need something new if you want it to work.
You don't need a specific ban. Just even the playing field for traditional finance. Either make cryptocurrencies follow normal banking/financial laws or make laws for traditional finance more liberal. Cryptocurrencies and especially PoW based ones are a terrible, slow, inefficient and vulnerable technology which wouldn't survive a day of fair competition.
Technologies being subject you the same rules.
The reason I can buy drugs, play online poker, request a ransom or just send money to an unknown at the other end of the world with crypto but can't with a bank transfer isn't technical. The only reason is that banks have to comply to regulation while crypto doesn't.
There are a lot of regulatory burden when trying to make a financial service. KYC, anti money laundering, reporting, privacy, data storage etc. There is basically none (that is enforced anyway) with crypto. That's the only reason it got traction. The technology is terrible. It's hard to come up with something more inefficient, expensive and hard to use. The one and only reason it's used for anything is that it evades regulatory burden.
It's literally "let's burn a lot of coal to run our steam powered car because steam powered cars are not subject to traffic laws" kind of situation. Fix and enforce the law and the steam car disappears. The same will happen to cryptocurrency.
PoW cryptocurrencies are still way more energy- and carbon-efficient than armored vans/trucks carrying valuables around, which is the other accepted method of decentralized, trustless value transfer. We shouldn't call for stuff to be "banned" that we can't even be bothered to understand properly.
Ah yes, I’ll never forget when I purchased my house and the armored truck delivered my money to the buyer. /s
Most transactions don’t need a decentralized trust-less value transfer. But bitcoin is not only attempting to displace the small sliver of transactions that DO need trustless/decentralized
I put a 20% downpayment on my house and it involved a wire transfer that was scheduled for days in the future that required my physical presence at a centralized authority (bank). With Bitcoin, that process takes less than an hour, with very little "fee", and doesn't involve any of the "trust", which was represented in my example by time delays, scheduling, and traveling.
Oddly enough making buying a house faster is not really a priority for me. I appreciated that mechanisms of trust kicked in for that transaction - my bank wouldn’t release mortgage money to anyone but a lawyer or notary, and my notary would not have released money to the seller unless they actually owned the place and transferred ownership to me.
Making that trustless means I am open to scam sellers, and have to trust the seller (who stands to gain if they can trick me) instead of now where i trust the bank and the conveyancers instead.
> Making that trustless means I am open to scam sellers
Cryptocurrencies support escrow transactions, where a mutually-trusted third party (e.g. an arbitrator) can decide whether the transaction should go forward if buyer and seller disagree. This is still "trustless" compared to other systems because it only depends on mutual trust wrt. each individual transaction, not on a pre-defined central authority.
Right, so my conveyancer (trusted by me, and my lender) escrows the money (lent to me by a bank that trusts me) via blockchain. How is this better than the current situation, where wire transfers are used, except 1) smaller fees 2) massive carbon footprint? Because I didn’t see any armored cars involved to offset the PoW energy.
I’m not saying blockchain is useless - just i see no use for it when it comes to the way that 99% of americans buy their house - and if 99% of americans switch to btc for this kind of transaction the carbon cost would be massive.
The carbon argument is so weird to me because if people understood blockchains it would actually bolster their argument.
Proof of Work Blockchains use the same energy whether there are any transactions or not. The idea that your participation or lack thereof deters demand in protest of the carbon footprint is just … wrong. Inaccurate.
Sure it’s not a direct per-transaction carbon footprint - but every transaction has mining fees, and more transactions mean there’s more money in mining which means more miners which means more carbon burnt.
The price on exchanges has little to do with onchain activity, which has always been a criticism of the valuation metrics, you are somehow reappropriating that retroactively in a way incompatible with the other criticism
its mind boggling… I would say more but I’m reserving my thoughts in case you have a rationale that is more convincing
I for one do not want to send hundreds of thousands of dollars to someone else without any kind of reversibility or fraud-protection guarantees. I'm willing to pay a token amount in terms of fee and delay in order to ensure that.
Poisoning the water supply to increase the margin on bottled water is a valid business plan. Banning that is a good answer, and neither authoritarian nor fascist.
We can all agree we want clean power, which is entirely in the hands of the suppliers. Rather than pushing one’s preferences around on users and forcing them to do your will, why not go after the Coal and Gas producers — the supply.
People get upset when you want to ban things because it’s a bully move. The bully thinks their opinion is true and correct and others must succumb to their will.
Your use of words like "bullying" and "pushing" suggests you frame this as some kind of antagonistic battle instead of arguing on the merits.
People don't get upset when we outlaw things because the rule of law is a great thing in general. People enjoy not getting mugged or having their water supply poisoned. Please tell me how a ban on poisoning the water supply is "bullying".
The point of contention is your claim that bans are "fascist". You have not substantiated that claim, so I don't really know why I even try to refute it, but I tried with a simple illustration that sometimes, banning things is not "fascist", and not bullying.
I agree that fossils fuels are being subsidized where they should be carbon taxed, and that's the only reason most crypto mining (and many other industries) using fossils fuels is still viable. But I disagree that bans are always authoritarian "madness". We ban things all the time. There are many substances and contraptions that you must hold a license to own or use. We have ban murder. We don't just disincentivize it, we actually do not allow it. And no one disagrees that this should be the case.
And this isn't a slippery slope argument, we are now at the point where irresponsible care for our climate is translating into actual lives lost.
Also, this is a perfect example of calling something "fascist" that categorically does not fit the definition of fascist. Fascism requires you to believe that one arbitrary group of people is superior to another, and that the govt/legislation should reflect this.
The arbitrary group is nevercoiners, who read something years ago and did some napkin math and drew some very wrong conclusions. They think they are better than the Bitcoiners. The nevercoiners are fascists.
I thought you said this was a war on general purpose computing. Wouldn't the arbitrary group then be anyone who uses a computer?
I don't want to restate everything I said above (which you definitely need to reread before you even consider responding to me), I just want you to see that you are a walking meme and should take a moment to reassess.
A living breathing walking meme. Are we not all? I’m not worried about this persona at all. It is purposefully as it should be. It services this role. We both agree you can’t stop people doing math.
PoW really is the tragedy of the commons epitomized. The only way it could be more on the nose is if mining literally required a random human being to die.
In 2019 the US used 18.27 billion gallons of fuel from airlines. If we limited flights to once a week we would save an amazing amount of fuel and increase flight occupancy, and decrease consumption as people would need to decide if the trip really needed a week layover if they couldn't take care of the problem in a 24 hour period. Much more than crypto consumption unless I am wrong.
Edit:
Commercial aviation makes up 918 million metric tons of carbon.
Your example shows the opposite of what you intend. Civil aviation has many positive effects. Bitcoin is questionable in this regard and yet is a non negligible fraction of a massively socially significant activity like flying
There is no need to ban them unless there is a particular local problem, like in Kazakhstan. Because proof of stake is technically superior for network capacity, proof of work blockchains will slowly die away. Users switch to better solutions. For example, there will be only proof of work coin (bitcoin) left in top 10 market cap by the end of this year.
Can you imagine if PoW cryptocurrencies hadn't of existed, that the climate change narrative would have been adapted to apply as "the world really needs to ban triple-A gaming" instead?
This story reminds me of an urban legend that the Sony Playstation 2 consoles almost hit a snag on an export ban because of "potential military applications". Different era, different context I suppose.
This will be an interesting experiment to see what happens when these companies spurn the customers that created the conditions for this market to exist. I hope Apple or another company is smart enough to use this opportunity to scoop up disenfranchised customers and let AMD and NVIDEA waste more time serving the Ponzi industrial complex.
Next they'll throttle internet speeds for miners like was proposed to slow down HFT, which now can place an order in the time it takes a photon to go 90 ft [2016 reference].
Just make the card the best it can be, meet the demand, and make money AMD, NVIDIA, etc.
Bad at crypto and at games if the specs are anything to go by. It's true the 6600 and 6700 series cards punch quite a bit above where you'd think they'd be relative to the 6800 parts but this one looks like an absolute bargain basement card.
Who knows, maybe AMD is going to pull another rabbit out of the hat? Seems unlikely with a 64-bit bus and 4gb vram though.
263 comments
[ 3.0 ms ] story [ 293 ms ] threadOf course, 4GB is typically considered the bare minimum for modern 1080p (let alone higher) gaming and this is just a budget card which was definitely not limited with any mining considerations.
Like the article says, it remains to be seen whether this will also make the card bad at gaming, but I think that AMD at least had a coherent plan here.
RDNA is far more efficient at gaming. GCN was compute-focused (higher TFLOPs, but weaker in practice). Case in point, micro-benchmarks show that VRAM latency is ~100 nanoseconds on RDNA, but ~300 nanoseconds on GCN.
RX 6500 XT is aimed at roughly the same specs as the RX480, but should perform slightly faster in practice, along with Raytracing support, Infinity Cache and therefore faster VRAM latency (probably faster fill-rates and other microbenchmarks important to gaming), all at 107W instead of 150W.
---------
As such, Rx 480 would be a better mining card, while 6500 XT would be a worse mining card, but better video game card.
Its still stupid for AMD to market their cards like this, but they're not lying... (and whatever argument marketing wants to make... well... that's their job. It all looks stupid to me but I can't honestly call them out on it since it is true...)
-------
Turns out that video-gamers who want more uniformity and less jitter between frames want higher-clock speeds (notice the 2000+ MHz clock speed on 6500 XT) rather than raw TFLOPs (Rx 480 ran a slow clock of ~1000MHz but had many more compute units / SIMD units computing in parallel).
Width of GPU leads to more variance, because its harder to load-balance your parallelism. If you have a narrower GPU (fewer compute units) at higher speeds, its easier to reach more consistent frame times / less jitter.
1. It sucks from big picture perspective, as we are "back where we were 5 years ago"
but
2. It's great from pragmatic perspective, as we are otherwise today in "Wish I could afford a card from 5 years ago" / "there's no such thing as a budget GPU" stage.
Do you not see the difference between GCN, RDNA, and RDNA 2?
At a minimum, this RDNA 2 6500 XT has raytracing cores that the two previous versions do not have. I'd assume that most video gamers care more about the raytracing feature (especially as more and more games are using it) than H265 encoding.
Remember: Raytracing instructions were invented on RDNA2. The 5500 doesn't have it, and the 480 doesn't have it.
-------
I mean, seriously. Would you rather have the existence of this card? Or the non-existence? If you want something faster, that's why the 6800 exists. This 6500 XT is for people who want a barely-gaming ready card at the cheapest possible price. If that's not your use case, then don't buy the card.
But every now and then, I build a computer for a child who doesn't play many video games, or for some uncle/aunt who doesn't know much about computers. I never like the idea of going iGPU only, because these people inevitably find at least one weird game that they like to play. So I like to put in a $100 to $200-class card in there. Not expensive enough to break the budget, but high enough that it'd perform decently on 10-year-old games and maybe a few indie-games.
Yes?
Minecraft Raytracing and Quake Raytracing are rather low-specs and probably would run on an RX 6500 XT. You might have to drop down to 720p but it probably would run.
They may have frankensteined the engines to support ray tracing but these are not modern games at all.
[citation needed]
Strictly speaking, it will "run", but if it run well that's another thing entirely.
It was the reason why the meme for heavy game morphed to "but can it run Minecraft RT?" for a while.
2x the clock speed is crazy.
First partners selling the part in Europe already have a hefty 50% surcharge [0].
[0] https://videocardz.com/newz/amds-199-radeon-rx-6500xt-offici...
PS5: 10.28 Teraflops, 36 Compute Units running at 2.23GHz (variable frequency)
XSX: 12.11 Teraflops, 52 Compute Units running at 1.825GHz (fixed)
Yes, this seems to be what's happening here. Less power used on memory bandwidth, means more is available for other things. This card is not so much "bad at mining" and more like trying to design for a different niche that might be underserved by all the "good at mining" GPU's.
But with the power consumption scaling with the square of the frequency, the difference in power draw can be absolutely significant. That's the reason many people (me among them) chose to underclock their 5700XTs
I'm playing around more with deep learning as of late. Are the same technical specifications one would look for in a 'miner-friendly' GPU the same ones I'd look for in a GPU for Neural-net computations?
> there isn’t an 8GB version of the 6500 XT. The 6500 XT also uses a 64-bit memory interface
It only comes with 4GB of RAM on a narrow bus which makes it useless for Ethereum mining. It doesn't seem like it uses the same "LHR" technique as Nvidia.
While I can see a justification for the lack of 4K H264 encode support (kind of pointless for a card at this level of performance), removing the AV1 decoder and H265/HEVC encoder is something I don't understand.
This makes the card completely useless for both game streaming and media creation for no discernible reason.
The holy grail would be a direct replacement backend that could be fed into TF, like CUDA.
ROCm has a direct replacement backend that can even take CUDA code (it's designed to be incredibly similar). It's called HIP. It's just that no one wants to support it. That is actually how TensorFlow on AMD works (mostly), and you can compile the latest stable release that way.
I can boot into Linux, and swap into Windows in 2 seconds with this setup. I have a dirty 20 line Bash script that deals with detaching the console, and passing the right things to the right place, but it all works.
ROCm on consumer cards does not work well. The tooling sucks. Massively. I don't understand why AMD doesn't have an extra team of 20 devs working just on the tooling.
Using DirectML with Windows Subsystem for Linux gives you better ML GPGPU support then AMDs native tooling.
That's a matter of opinion I suppose, but I don't personally find that passable.
>Using DirectML with Windows Subsystem for Linux gives you better ML GPGPU support then AMDs native tooling.
DirectML sucks even more than ROCm, IMO. Also, WSL sucks more than a normal VM.
What is a setup that would be passable then? I think a setup like the one that I have described [I believe] would be impossible with Hyper-V or ESXi (though, not that I have even attempted it with either).
Compare this to Nvidia or even Intel where I can run GPGPU API's (CUDA, oneAPI) on low-end, consumer-grade hardware.
The problem, ultimately, is that AMD does not even care.
I agree that AMD is dropping the ball.
CUDA support (or rather lack thereof, which isn't AMD's fault) plays a major role w.r.t. software support, but AMD's compute architecture isn't DL-focused either.
The MI250X compute part for example has a FP16 to FP32 ratio of 8:1 and a FP32 to FP64 ratio of 1:1; in other words it's an absolute beast at GPGPU compute.
The 6900XT on the other hand has a FP16 to FP32 ratio of just 2:1 and a FP32 to FP64 ratio of 1:16 (i.e. it's severely restricted at high precision workloads and OK at half-precision).
Comparing this to the specs of NVIDIA cards, they're still vastly superior on paper. The consumer versions of Ampere only get 1:1 (FP16:FP32) and 1:64(!!! FP32:FP64) respectively. But then again, NVIDIA cards feature dedicated "tensor cores", which have no equivalent on AMD consumer grade hardware.
The main selling point for NVIDIA, however, is software support and mindshare. They started to buy themselves into academia in the late 2000s by sponsoring labs and providing a vast ecosystem of software libraries for deep learning and GPGPU support in general. This not only helped kickstarting the deep learning revolution but also tied their hardware and brand name to GPGPU, which basically became synonymous with CUDA at that point.
Most other software focused exclusively on CUDA, so GPGPU on AMD cards is not well supported despite the potential.
Blender is a popular example for this. After discontinuing cross-vendor OpenCL support, which effectively limited GPU acceleration to NVIDIA cards, they added support for AMD cards in the latest version. Even then, only the latest generation of AMD cards is supported for some reason.
Other renderers like OctaneRender still only support CUDA. The situation is just as bleak in video editing software, were major companies like Adobe only support NVIDA and Intel (at least on Windows) or have poor OpenCL support (which users then blame on AMD of course).
There is some hope that software support for GPGPU without CUDA (maybe using Vulkan Compute?) will improve later this year with Intel (re-)entering the discrete desktop GPU market.
At least, that's what the first graph I found said.
so Bitcoin crashing, without Ethereum existing, wouldn't mean anything. thats how I separate the two without it seeming pedantic to me.
(I recognize that there have been issues with Wayland support in the past, it creates issues for some developers, and support for modern Nvidia cards on most other operating systems is non-existent. There are likely other issues, but it seems like most of the issues would affect a small number of people.)
Who are you to dictate what someone does with energy they pay for? Just because you do not like what they do, doesn't mean you can ban it if it's not harming you.
There is no evidence that Bitcoin's usage of renewables is preventing others from paying for renewables too.
If everybody is using renewable energy exclusivity for everything, is still wastage. PoW only work by adjusting it's consumption upwards to create scarcity, e.g. one block per 10 minutes. At that point we will be useless solar panels, covering land uselessly, etc, etc. Bitcoin is arbitrary scalable wastage to create arbitrary, artificial scarcity.
The point is: Energy wasted is energy wasted. Bitcoin will eat everything till the point where some external factor prevents further expansion, e.g. how much are we willing to hurt the environment and ourselfs in exchange of the block reward. The person with the highest threshold wins.
It's not really not complicated. This is literally the last scene of the hitchiker, where they plan to kill all trees, so they can use the leaves for currency. The success of bitcoin makes be loose all hope in humanity.
Authoritarianism isn't an ideal outcome.
This is also in keeping with tons of existing regulations: you can't buy a monitor which uses 2,000W because various governments have caps for how inefficient you're allowed to be. That's far from perfect — or sufficient to deal with the climate change crisis — but it's not like it's some sort of unprecedented imposition on personal freedoms to cap inefficiency.
so you made a judgement on what an activity is useful. There's no such objective judgement possible, since usefulness is a subjective matter.
Therefore, pricing carbon correctly (and thus pricing electricity correctly) is the way to provide all parties the full information on cost, and each would decide how best to use their own money to buy energy.
> so you made a judgement on what an activity is useful.
No, I understand how the system works. Bitcoin uses power as a function of the aggregate network capacity — it must to avoid a 51% attack — and that means that the power usage not coupled to the number of transactions being processed: even if Bank of America wasn't already many orders of magnitude cheaper, they wouldn't buy a mainframe they didn't have work for just to stay tied with Wells Fargo. It's the computational equivalent of Luftansa flying all of those thousands of empty flights.
> Therefore, pricing carbon correctly (and thus pricing electricity correctly) is the way to provide all parties the full information on cost, and each would decide how best to use their own money to buy energy.
This would help considerably but it's not enough: we need to radically pivot the global energy system and that will involve things like shifting where major power consumption occurs (e.g. ideally things like blast furnaces or smelters would be located near sources of hydro, geothermal, nuclear, etc. power) and getting everyone to convert to things like EVs en masse.
Unfortunately those decisions take a longer timeframe than shipping a container full of cryptocurrency mining hardware to wherever the current cheapest source of power is. That adds a lot of noise to the pricing signals we need to drive business decisions, especially for major capital expenditures which require years of planning. It doesn't help if, say, Iceland adds 10GW of clean geothermal power but the price stays the same or even goes up because a bunch of miners showed up the week after it went online.
You didn't read or you didn't understand what I said. It's waste because it's designed to create a scarcity. Not by the judgement of anyone of the usefullness. The power bitcoin uses is strictly a function of how much power is available, not what service is provides. Double the number of transaction rate: No change in power usage. Half the price of electricity: Double the power usage. This strange lack of connection between utility and resource usage makes it wastage. It might be hard to grasp, because there is nothing unlike it.
> Maybe we should closely audit your energy usage and ban those activities we find objectionable.
As stated above the energy consumption of bitcoin is not bad because the a decentralised ledger is objectionable, but because it's a controller (in the sense of control theory) tuned to make energy a scare resource. This is a scheme equivalent to poison every aquifer in the world to profit from the scarcity of water. It's doesn't matter if scarcity of water is useful for some apparatus, or if the task the apparatus performs is objectionable. If you want to monitor my non bitcoin energy-usage demonstrates you didn't understand is difference between utilisation of water for the sake of using it and for the sake of wasting it. Usage can be made more efficient, and new resources can cure a lack of water. In the case of wastage the wastage can by definion not become more efficient and new resources will not make it last any longer, but simply call for an increase the rate of poisoning.
Even if bitcoin had any use case besides randomware, which it doesn't, even if it did solve every problem the financial world had, it would see be wasteful. While most other technology becomes more efficient with further use cases, PoW becomes more wastefull, for reasons state above. The only way to make it more efficient is to make energy less aviable to humanity, as such are the perverse incentives of this scheme.
IMO this is the strongest case for a carbon tax - it respects individual freedom.
You want to cruise in your diesel-sucking massive launch? You want to take your personal 747 out for a joy ride to the Bahamas?
Fill your boots, do what you want without any guilt - your environmental impact has been offset by your higher fuel bills.
Proof-of-waste systems like Bitcoin are uniquely inefficient due to the way the network adjusts to increase the difficulty — an aluminum smelter uses a ton of power but they don't build new ones in excess of market demand for aluminum — and they're easy to ship around the world. That means that demand always goes up and any time some place has cheaper power for a non-trivial period of time someone will probably ship a bunch of mining hardware there.
This is a problem because right now we need the entire global economy to electrify as soon as possible. That means things like industrial processes need to be relocated to cheap sources of clean electricity because you can't just plunk a coal power plant next door, etc. and there are also challenges with peak power demand requiring upgrades to electrical grids or other equipment. Having so much power dedicated to something which has never found a useful purpose is a source of friction we don't have time for.
A member in a democratic society that doesn't give you a fundamental right to mine bitcoins.
> Just because you do not like what they do, doesn't mean you can ban it if it's not harming you.
Believe it or not, this is not how society works.
> Believe it or not, this is not how society works.
and yet, we are now unbanning a lot of activities that some people dislike - such as smoking pot. the OP's point is that not everyone is in consensus about the energy use of PoW being a waste.
This is an example of Jevon's paradox, namely that increases in efficiency or quality are offset by increases in demand. It's one of the reasons why we have globally barely made a dent when it comes to the increase of renewables as a total share of energy consumed despite the fact that the capacity has increased pretty much exponentially.
This is made even worse by the fact that green infrastructure itself necessarily often consumes non-green sources, in particular during construction.
This is not true. Often, renewables aren't available to some locations but abundantly available in others. Other times, some people buy electricity from renewable providers and others don't.
With 50-75% of Bitcoin mining being renewable, one thing is clear - it's helped the adoption of renewables and made them cheaper.
Any excess capacity spent on crypto currency is waste. It could go to scientific computing projects, it could simply not be used. The chips for the equipment (we are in a global chip shortage) could go to something that actually contributes to the real economy.
We just need to increase carbon taxes at the same rate that renewables become affordable.
We have about a few thousand pressing issues that spur us on to push for renewables as it is, we don't need to waste energy on purpose.
Crypto mining is one of those issues, we don't have a choice. Your argument reduces to "if we had no issues pushing us, we would have enough issues pushing us" which is a contradiction.
Miners are chasing profit margin so they'll use whatever's cheapest. If that's renewable where some of them are, so be it. And if everyone else start reducing use of coal and that becomes cheap, miners who exploits that to derive higher profit margin will grow faster I imagine.
I would attribute growth and economies of scale that renewables have achieved to renewable proponents, industries, government subsidies and loan prorams. Crypto miners are energy consumers and some of them use renewable energy. But how do we go from there to miners being some kind of force for good in this context?
That is kinda the main point of a carbon tax... it is to make carbon usage LESS efficient, since you have to pay an extra tax in addition to the cost of the energy itself. Ideally, the money gained from the carbon tax is then redistributed to everyone equally, and that money can then be used for whatever you want.
i dont think that's a good idea - the redistribution does not decrease carbon.
I think the correct carbon tax is the amount of money it takes to fully remove the carbon emitted (at current tech levels). This pot of money is then used to actually remove carbon.
A pigovian tax by insurance would work better, I definitely think environmental protection is needed but politicians will just use the money they get in nefarious ways.
Collection can be simple too. Charge at major sources like coal mines and oil wells, and the fee gets rolled into the prices of everything downstream.
In fact, estimates peg Bitcoin's use of renewable energy to be between 50-75%. This is a fact lost on HN, which hates crypto at all costs.
And crypto needs to be banned because it's all a gigantic scam that has a gigantic environmental cost attached to it.
And my emissions are offset for the fee that I pay.
Carbon credits, more planted trees, whatever. My net impact is likely lower than anyone that doesn't pay for renewables, even with mining.
> And crypto needs to be banned because it's all a gigantic scam that has a gigantic environmental cost attached to it
For PoW, it's the purest form of currency - the value of energy itself.
For other currencies, it's no different than precious metals.
Trust HN to call decentralized technologies enabling pseudonymity/anonymity scams on one hand while advocating for decentralized technologies enabling pseudonymity/anonymity on the other hand.
Also, from my perspective, you should stop eating meat. Your eating of meat provides you nothing but selfish pleasure while causing a ton of carbon emissions - and actual tangible suffering on scales never before seen on this planet. The magnitude of suffering caused is tens billions of times worse than crypto. Let's ban meat first.
You might want to revise your argument if you're not going to acknowledge the fact that meat has a good number of calories per kg.
Is it worse than, let's say, Christmas in this regard?
The design of Bitcoin ensures that does _not_ happen there, and it's currently using a massive amount of power to process a tiny fraction of the total economy, especially after you account for the transactions which are not linked to real economic value (money laundering, wash trading, etc.). The design ensures that if someone did find a useful application which saw popular adoption, usage would need to be many orders of magnitude higher than it is now.
And banning crypto mining is impossible. It's arbitrary math. You can't ban a computer from doing arbitrary math. At best you could discourage businesses taking crypto as payment, but even then people can operate 100% in crypto and avoid any regulations.
That's not how it works. It just means that instead of everyone using 10% renewable energy you use 100% and 9 other people use 0%.
Now if PoW was using 10% of the world power maybe it'd worth it but I just don't see how you can justify such a ban but only apply it to this one thing.
https://www.prima.co.uk/fashion-and-beauty/fashion-tips/a269...
People are usually opposed to line drying clothes because of aesthetics and feeling like it makes their neighborhood look “poor”. Luckily quite a few states have passed laws making it illegal for landlords or HOAs to ban putting up clotheslines.
And the fires!
I'm also shocked that people do laundry so often. Not sure why people are doing more than one or two loads per week and that's a problem the government could look at tackling (not so much with PoW coins).
[1] https://www.nytimes.com/interactive/2021/09/03/climate/bitco...
[2] https://www.nrdc.org/sites/default/files/ene_14060901a.pdf
We do about 4 loads per week here. I hope that the newer heat pump based dryers are reliable!
It's absolutely crazy. I know that there's a big lobby behind mining, I absolutely respect what they have being doing protecting Bitcoin i.e. and making the project survive... even if it somehow completely lost it's original purpose. But FFS guys what's the purpose of having cryptocurrencies if the cost of that is aiming to destroy the world, or much more probably, to win a last minute complete cryptocurrency ban from all governments with the handy climate change as an excuse?
[1] https://www.thebalance.com/how-much-power-does-the-bitcoin-n...
They don't really require more power, they are just getting more valuable so it's been profitable for more people to increase their mining investments. The limit is that they are not infinitely valuable, and if anything mining rewards are decreasing.
PoW is wasteful by design. The more hash power you throw at the network, the more difficult the PoW becomes. Any efficiency gains necessarily have to go towards more hashing, instead of maintaining the same level of work at a lower power draw.
A gaming computer at ~400W, left on 24/7, is 67,200Wh. That assumes you are playing 24/7 - your idle wattage in sleep is < 100W.
If you run your dryer every night of the week for an hour, you are using about 38,400Wh.
You can game 24/7 and dry your clothes every night of the week and STILL use less power than a modern, efficient single miner node. About 50% less!
Bitcoin is a huge waste of electricity, and it is WAY worse than anything else you listed.
[1] - https://www.thebalance.com/how-much-power-does-the-bitcoin-n...
This is what i dislike about the argument that PoW should be banned - that it wastes energy that could otherwise have been used for something else. I argue that it is not a waste, because the person who is paying for the energy does not consider it a waste. No one else can make this judgement.
The only weakness of this argument is that some energy sources have externalized the cost of emissions. I would very much like to see an emissions tax to account for it (which makes sense to do, regardless of one's position on PoW).
However, to say it's all a waste is just an emotional argument because you don't like it. That's fine but it's opinion, not objective fact.
I don't doubt that bitcoin miners use a lot of energy, but I hesitate to trust a source that doesn't know the difference between power and energy. Also, they're lying about the energy consumption of Sweden, which actually uses 645.7 TW/yr as of 2012 [0]. They number they quote is actually the electricity consumption of Sweden [1], which is like 5x lower.
[0]: https://www.wolframalpha.com/input/?i=energy+consumption+of+...
[1]: https://www.wolframalpha.com/input/?i=energy+consumption+of+...
Who will make a decision as to what is considered frivolous.
Not to search very far, I consider trucks in US for non-contractor use to be frivolous, but I don't go around telling people those need to be banned.
No typical business is going to work with BTC addresses on the raw. They're going to (and currently) using simple APIs like Coinbase API (https://developers.coinbase.com/) to "implement BTC receipts".
As far as US business is concerned, the typical payment processor for bitcoin is Coinbase. They handle the bitcoint transfer, transaction history, and conversion into USD all at the same time.
Bitcoin works the way it is intended, that is peer to peer payments. Everything else is just ecosystem, which can be replaced. Coinbase is important as an on-ramp for users for now, yes, but their importance to the network is peripheral.
There's your theory, and there's the practice of the internet and business. In theory, they should be the same. In practice, they're different.
I’m talking about Bitcoin, and you better believe I run a fullnode. No provider required. I recommend you run a Bitcoin full Node too, and transact without any 3rd parties!
The users/miners decided, in decentralized fashion by contributing resources. Decentralized!
Governments have this general monopoly on the use of force, which is why they tell you you can't do things like sell heroin. If you disobey them, they will send men with firearms and make you stop.
Governments can obtain their consent in a bunch of different models and lots of people argue about which ones are better.
A "world wide ban on crypto" probably is not going to happen but more and more countries are banning it. I suspect many more will continue.
Still, I will address your comment in good faith.
It is not a checkmate. I am simply just trying to see what world model you have in your head. We clearly disagree, but it does not mean we can't openly discuss those disagreements. It is mostly the beauty of HN. If you present a logical argument, I might simply say, it makes sense. Right now I am confused.
Are you officially arguing government is the one to say as to whether technology is banned? If so, which government speaks for humanity as a whole?
<<This is the equivalent of screaming you're a sovereign citizen when the police ask for your name.
Is it? Can you elaborate a little bit? The analogy seems a little off.
<<Governments have this general monopoly on the use of force, which is why they tell you you can't do things like sell heroin.
Yes. Government monopoly on power has been a well established theory. We are not breaking new ground here. Is crypto heroin in this scenario?
<<If you disobey them, they will send men with firearms and make you stop.
Yes. See above. How does that help your argument?
<<Governments can obtain their consent in a bunch of different models and lots of people argue about which ones are better.
Yes. See above. How does that help your argument?
<<A "world wide ban on crypto" probably is not going to happen but more and more countries are banning it. I suspect many more will continue.
Maybe yes. Maybe not. It is not, however, an argument for or against crypto banning, whether it should be banned or not. It is not an argument at all. At best, it is just a speculation.
The answer is the government.
Governments are restricted to their borders but they do like to cooperate, and so EU may pass a general crypto ban and the EU may even ask the nations it trades with or that want to join to do similar.
But there is no real question that a government may ban something in it's borders.
"Ah, but I can get away with crime sometimes! Thus the government ban doesn't matter!" is the usual response. And sure, some small crime goes unpunished, but in general people don't want to overtly engage in criminal behavior. No business would touch crypto, which harms it's value and use.
Should we ban crypto?
Well, we can explore why countries have. The answer is a mixture of "it makes tax evasion stupidly easy" and "it uses too much power". With parts of the world paying INSANE prices for fuel right now, a lot more governments will latch onto that second line of reasoning. In general, governments detest tax evasion and things which aide it.
We do in fact generally like tax revenue. We like roads and schools and such, and most people are opposed to tax cheats. Nobody particularly cares for high electricity prices or polluters either.
"Well, what about those crypto operations which pay their taxes and are funded by renewables?" Sure, perhaps those are allowed to exist, but that comes with two problems. 1. It dramatically shrinks the mining pool, making sybil attacks easier. 2. It reduces processors which means less businesses will accept crypto.
Crypto already has a strong association with open air drug markets, and for good reason. As legitimate usage declines, the desire to ban a tool used primarily for crime will increase as it should.
Crypto is already being banned and it probably should be banned. If not banned then regulated, but regulation would harm it significantly.
I take issue with the statement "it makes tax evasion stupidly easy". The moment you touch any US-based exchange, your information is up for grabs by a variety of agencies from IRS to various LEOs. It may make doing the deed easy, but it does not really change much. It has been years now since blockchain made it into mainstream. There are companies specializing in selling blockchain insights. Unless we are not talking US. But US has an entire cottage industry devoted to making sure that the wealthy pay as little as possible. I am not sure banning crypto would help here. Killing lawyers and accountants could ( I jest.. it is just a reference to Henry VI ), but that is about it.
<<"it uses too much power"
I find this argument amusing. I am paying for it. If you don't like my use, either raise the price or, well, leave me alone. But therein lies a problem, energy prices are not an equilibrium point and people do not want to give up THEIR toys, but an appropriate scapegoat will let them keep their toys and feel righteous indignation. It is a win/win.
<<The answer is the government.
And this is where it gets fun as correctly noticed. Even without considering the distributed nature of the technology itself ( because there are way to circumvent that as internet splintering has shown ), government these days may, if not will, disagree for a variety of reasons including but not limited to "they are for A, therefore I am for not A", making it a protracted game of whack-a-mole ala piracy, war on drugs and multiple other endeavours governments proved to be so good at. To sum up, you may reasonably argue they have the authority, but do you reasonably think, they will actually pull it off?
Residential miners are using subsidized electricity. We do that so Grandma can afford to be warm and run her dryer.
If we forced miners to use commercial electricity plans then they would in fact pay a whole mess more.
Still, let us say that I accept your argument, some miners actually bought various electricity producing entities. Now what? Are you going to argue they can't use the electricity they generate or are you going to argue that we can then force them out of business some other way. Either undermines some tenets of US society, but I am curious now.
The gov't is supposed to represent the people's will. So are you saying that it is the will of the people to ban crypto? Aka, is crypto something that is causing harm to society, like heroin addiction etc?
Let's go back to the prohibition era - why did the gov't decide to ban alcohol? And why did they re-instate it back afterwards?
I argue that gov't is not an "authority" - they merely execute the ban. The authority lies with the average views of society. Bans on slavery, for example, is only possible because most people slowly came around to disliking it (mostly people who don't benefit from it - while those who do vehemently fought against the banning).
Not to mention, the US forced the anti drug ideology on the rest of the planet through economic coercion
So yeah, you are not wrong about the general framework, but do you think drafting such a law right now would be..hmm.. palatable to the general populace. After all, consent of the governed is a thing.
And as for crypto specifically.. well.. the time to squash it has long passed. There is people with real money in it now. The best you can hope for is regulation.
What does you not liking trucks have to do with the issue of putting limitations on mining crypto?
Just because you don't care about this random issue doesn't invalidate serious concerns other people have.
Also, this is a pretty bad faith argument to say 'who can make these decisions'. We make rules in society all the time. If enough people can be convinced for a particular regulation and the laws of the society allow it, then it'll be implemented.
At the end of the day, trucks are there to illustrate a point that people may not agree what is considered 'good','efficient' or whatever qualifier you want to add. Trucks - those humongous behemoths intended for largely for contractors, but bought people with more money than sense for reasons that are not 'good','efficient' or whatever other qualifier you would want to add to crypto - are largely inefficient for a family unit. Some could easily argue that those should be banned for personal use as their gas-guzzling abilities are a MUCH greater threat to climate change than the entirety of crypto mining operation.
But do I suggest people are mandated to only get one Prius every 5 years? No. Do you know why, because the concerns of the people may be real, but we have not, as a society, yet decided that it is ok for those concerns to override individual choices.
You are right about society. You are wrong about the argument.
I agree with OP though, we're within an order of magnitude of other legit uses, so let people be free. Increase electricity costs or taxes if you want.
I think OP point still stands. This is comparable in term of "waste".
[0] Siemens WT47R400: 176 kWh / year for 160 cycles by energy label, 1.4 kWh for a full run.
But the main difference is that whatever you compare with those blockchains, it will never have the critical flaw: the automatically increasing difficulty. Your dryer doesn't periodically flood the drum to make it harder because more people are interested in tumble drying.
However, market forces guarantee that the difficulty will go up to the point where it's infeasible to add more hashrate, e.g. because it becomes uneconomical or there's simply no more hardware available. PoW workchains must work at maximum inefficiency. (This bears out in the data, too, I think. BTC had its highest hashrate at the beginning of this month, when energy price were at an all-time high and BTC was at the lowest over 12 months)
Classic whataboutism. First of all your entire comparison makes no sense, you have to compare alternatives for use. What's the alternative to a dryer. Using a line and hanging your clothes. Fair enough, what about in the winter in New York? A ventless dryer? Doesn't use substantially less energy. And so forth.
For every conceivable use case there exists alternatives that use orders of magnitude less energy. NFTs, sending money, federated computing, you name it, something already exists.
The only reason blockchain is a thing is because cryptocurrencies are booming in terms of how much they're sold for. The end.
I dry my clothes on a line in the winter in New York. It's below zero outside, but like 80 degrees and 0% humidity inside my apartment. Clothes dry fast. (And the space was already being heated, because I live in it, and it's against the law for the landlord to allow the temperature to go below 55 degrees, and they are paranoid, so it's more like 80-90 all winter.)
The heat is on a timer so it's not like this all the time, but I have to wonder what a world where thermostats have been invented would be like.
Large swathes of Europe have winters as bad or even worse than New York's. I've grown up in one of those countries and had never seen a home clothes dryer before visiting the US. I've also almost never seen them in London or Berlin where I lived after that and those have at most marginally warmer winters than New York.
Electrical resistive heating clothes dryers are an enormous energy hog.
Heat pump water heaters, though, are supposedly much better than resistive.
Of course, those taxes will probably just disappear into general tax revenues and wind up paying for other things, but it would encourage crypto miners to pursue green energy sources and, perhaps, find better uses for obsolete graphics cards.
why should crypto be specifically targeted for an energy use tax (ala, a carbon tax but with extra steps), instead of a general tax on energy use (aka, just implement the aforementioned carbon tax)?
We all want power to be cleanly generated. This demand side bullying needs to stop. Give us clean energy!
If I want to use a machine to dry my clothes, I pay for that electricity. This puts a limit on the amount of electricity I can consume to do it.
Cryptocurrency is different. The more electricity you burn, the more money you make. I know this first hand - the computer I'm typing this on is running gminer to mine ETH on flexpool.io with my RTX 3070 card. I've mined $2,800 of ETH in the past 10 months.
Something that most people don't understand is that there is a finite amount of ETH that can be mined per day. Hashrate determines the size of the slice of that pie you get. Therefore the incentive is to consume infinite electricity! Of course miners can't do that because mining gear (i.e. GPUs, ASICs, FPGAs, whatever) is hard to purchase and electricity isn't infinite.
But that doesn't stop miners from trying. I know people that do anything they can to acquire GPUs - they have bots combing the Internet, they buy whole systems that have GPUs, rip out the GPU and sell the system, they've signed up for every GPU manufacturer's wait list, they stalk Microcenter, whatever they can do. They seek out areas with cheap or even subsidized electricity and latch onto it like a parasite. It is nothing like anything else you spoke of.
there will be a new equilibrium - the more electricity you use, the less of it is available, and thus the price would rise to match the demand. This means there's a limit to how much you can earn (unfortunately you wont know this limit until you reach it, because the full demand for crypto is unknown atm).
But high prices of electricity would spur developments in energy production.
In any case, energy use is energy use, regardless of the purpose. As long as the person using it is paying for it, then it is a valid use.
I cannot understand how this is not a bigger issue being talked about in the media. Right now in the UK we are going through an energy crisis where people can’t afford to heat their own homes (I’m not blaming crypto for this). It’s absolutely maddening that in a world where climate change and energy scarcity is such an issue this would be a controversial topic.
My dishwasher is rated based on specific cycles, but I'm legally allowed to use it as frivolously as I want. I can run an A+++ rated dryer with a single sock in and send the photo to the police.
We don't limit energy use but inefficiencies. The equivalent in crypto would be to rate hardware by hashes per joule.
Crypto things are different because the incentives are different. I'm incentivised to use my appliances efficiently as the energy cost is just a cost.
On the other hand with PoW, I'm incentivised to mine efficiently but as much as possible because I can buy a thing that turns electricity into cash.
Fine, yes. Ban both.
There are a lot of regulatory burden when trying to make a financial service. KYC, anti money laundering, reporting, privacy, data storage etc. There is basically none (that is enforced anyway) with crypto. That's the only reason it got traction. The technology is terrible. It's hard to come up with something more inefficient, expensive and hard to use. The one and only reason it's used for anything is that it evades regulatory burden.
It's literally "let's burn a lot of coal to run our steam powered car because steam powered cars are not subject to traffic laws" kind of situation. Fix and enforce the law and the steam car disappears. The same will happen to cryptocurrency.
Most transactions don’t need a decentralized trust-less value transfer. But bitcoin is not only attempting to displace the small sliver of transactions that DO need trustless/decentralized
Making that trustless means I am open to scam sellers, and have to trust the seller (who stands to gain if they can trick me) instead of now where i trust the bank and the conveyancers instead.
Cryptocurrencies support escrow transactions, where a mutually-trusted third party (e.g. an arbitrator) can decide whether the transaction should go forward if buyer and seller disagree. This is still "trustless" compared to other systems because it only depends on mutual trust wrt. each individual transaction, not on a pre-defined central authority.
I’m not saying blockchain is useless - just i see no use for it when it comes to the way that 99% of americans buy their house - and if 99% of americans switch to btc for this kind of transaction the carbon cost would be massive.
Proof of Work Blockchains use the same energy whether there are any transactions or not. The idea that your participation or lack thereof deters demand in protest of the carbon footprint is just … wrong. Inaccurate.
its mind boggling… I would say more but I’m reserving my thoughts in case you have a rationale that is more convincing
This is an attack on general purpose computing. You cannot tell people what they can and can’t do with their machines.
You can try, but the smart ones will tell you to go f yourself.
Deal with the supply side costs, and let people do whatever math they like.
The analogy would be you telling people they have to drink that water a certain way and banned from drinking it the way they want. That’s fascism.
Everybody that wants to ban PoW is an authoritarian fascist. Might be an eco fascist, might be a boot licking state loving fascist… still a fascist.
No one is banned from drinking poisoned water. What is banned is poisoning other people's water. Really, this isn't that difficult.
If you want to make a point about supply, please make a point about supply instead of just repeating the word randomly.
People get upset when you want to ban things because it’s a bully move. The bully thinks their opinion is true and correct and others must succumb to their will.
People don't get upset when we outlaw things because the rule of law is a great thing in general. People enjoy not getting mugged or having their water supply poisoned. Please tell me how a ban on poisoning the water supply is "bullying".
And this isn't a slippery slope argument, we are now at the point where irresponsible care for our climate is translating into actual lives lost.
Also, this is a perfect example of calling something "fascist" that categorically does not fit the definition of fascist. Fascism requires you to believe that one arbitrary group of people is superior to another, and that the govt/legislation should reflect this.
I don't want to restate everything I said above (which you definitely need to reread before you even consider responding to me), I just want you to see that you are a walking meme and should take a moment to reassess.
Edit:
Commercial aviation makes up 918 million metric tons of carbon.
Bitcoin mining makes up 22 million metric tons.
This story reminds me of an urban legend that the Sony Playstation 2 consoles almost hit a snag on an export ban because of "potential military applications". Different era, different context I suppose.
https://www.pcmag.com/news/20-years-later-how-concerns-about...
Yes, I would be worried that for certain games; performance will be oddly super bad.
Just make the card the best it can be, meet the demand, and make money AMD, NVIDIA, etc.
Who knows, maybe AMD is going to pull another rabbit out of the hat? Seems unlikely with a 64-bit bus and 4gb vram though.