Didn't watch the video, but a quick point on how the energy efficiency of this works out:
Electric heaters are 100% efficient at turning electrical energy into heat. All the energy they take in, that's where it goes.
Bitcoin mining (and computers in general) are basically in the same boat. There's some ancillary energy use for blowing air around or shining lights or making sound waves, but for the most part the electricity turns into heat and does some math along the way.
Sounds great! Perfectly efficient! But the important caveat is that heat pumps are (effectively) more than 100% efficient. They can take 100 watt of electricity to transfer 300 watts of heat from outside to inside.
It's not technically a measure of "efficiency", we call this the "coefficient of performance" instead. Or we call it the "seasonal energy efficiency ratio", which is the same thing except multiplied by some constant factor because somebody wanted a scale that went up to higher numbers. Or for heating, the heating season performance factor, which I forget the specifics of, but is geared toward how well it works pumping heat in from outdoors when your heat source (outdoors) is very cold.
In short, it's just as efficient as any other electric heater, but heat pumps are 3-4x better than that.
>heat pumps are (effectively) more than 100% efficient
The CoP of a heat pump is greatly dependent on the difference in temperature potential of the source and sink. If it's extremely cold outside, it's very hard to pump heat from colder air into a house. When the temperature potential between the two reservoirs is equal, sure the CoP beats out electric heating but when it's high (When your house is warm and the outside is below freezing) its extremely poor.
Cool data. Appears the CoP vs outdoor temp plots are on page 14. Since the data doesn't provide temperature delta between the source and sink, you'd need to assume that the pump met the heat delivery requirements.
This is what we are doing at Qarnot Computing for 5years.
We created a heater using CPU as a heat source.
People send a docker image (3d rendering, financial analysis...) and we run it on heaters deployed in buildings in Paris where people are heated for free (we refund the electricity cost)
This is silly, CPUs are 100% efficient at converting electrical energy into heat. Where do you think that 'ancillary energy' is going?
Obviously heat pumps are more efficient - that's kinda beside the point though.
The main point here is you can take anywhere you had a resistive heating element and replace it with a mining chip, and you're now doing something useful with that energy you were turning into heat anyway.
Yes heat pumps are more efficient but there are plenty of places in your house you have plain old heat elements that aren't going away (water heater).
>Obviously heat pumps are more efficient - that's kinda beside the point though.
Only obvious if you know what a heat pump is!
Plenty of people will hear "Electrical heating is 100% efficient" and interpret that to mean it's the ideal solution, when in a lot of cases it's just about the worst.
But yes, if you have electrical heating, you could throw in some CPUs or ASICs instead and they'll work just as well. The return on investment for cryptocurrency may drop off a cliff when the next silicon process comes out and mines more for the same energy, it's hard to see people wanting to keep upgrading their baseboard heaters on the same cycle that computers get replaced.
Whether it's money positive vs resistive heating will depend on A) price of your electronics relative to a cheap electric heater, B) how it stacks up vs future electronics if the payoff doesn't happen while it's new, and C) how insane the cryptocurrency markets get.
At $430/month for the bitcoins, I think they can afford to keep the systems current. Obviously market conditions can change drastically, but they're just as likely to go up as down.
It's resistive heating that does something that most resistive heating doesn't do: generate some level of additional income in the form of cryptocurrency.
Only if you need full power heat 24/7 365. If you ever want to turn it off then it's costing you deprecation on very expensive hardware, vs a simple heating element.
Remember, few things need constant heat output 24/7 365, otherwise your going to want to generate some or all of your heat from other sources.
CPU's can't tolerate high temperatures so it only works for a small subset of things you need a heating element for.
Aka, you can uses them to cook with. Also, while you can use electricity to heat water it's vastly more efficient to burn natural gas to make heat than burn natural gas to make electricity to make it, ditto solar power etc.
They probably don't make as much sense in cold climates, but where I live, air conditioning in the summer (and spring and fall) dwarfs the energy usage from heating in the winter. So a device that makes it colder on average indoors would be a net win even if it didn't have the function of heating your water.
When I looked into these units, the potential repair costs over time led me to choose a highly insulated electric (Rheem Marathon series). The major parts like the compressor were not available through Grainger or even an HVAC supply store, nor could I get my hands on a retail parts price list. So I had concerns over vendor lock-in, and the initial capital premium would get vastly inflated in opex costs over time.
Has anyone had 7+ years experience with these, and if so, can you please share your ownership experience and the model & brand you used?
and you're now doing something useful with that energy you were turning into heat anyway.
Is it useful? My (limited) understanding is that most of the work is just arbitrarily difficult number guessing for the sole purpose of being throttled...
It's useful because it's useless. Different definitions of "useful" at play there. Proof of work is by design unable to do anything other than provide a probabilistic proof of energy expenditure, but such proofs are themselves useful.
That's true, I'm speaking more generally about bitcoin mining for heat, but it may make more sense in Siberia if electric heaters are the dominant heating method.
There are other reasons electrical heating isn't great, heat pumps aside. An average coal plant (in the US in 2015, I'm using EIA numbers) was about 34% efficient in turning the energy in coal into electricity. Then you lose about 5% of that in transmission, so from the heat released by burning coal you're only getting 32% of that as electricity at your house. Converting 100% of that into heat is pretty unimpressive in the big picture, you'd get 3x more heat by buying some coal and setting it on fire.
Obviously you can't do that, but a high efficiency gas-fired furnace will get 90%-98% of the combustion heat into your ducts, with the much smaller remainder going up the chimney. Unless you're on solar power (or some other renewable), burning fossil fuels on-site does better than resistive electric heating.
Of course that may also not apply to Siberia. If the ground is frozen all the time it's harder to run gas lines compared to putting in an electric grid. Mostly I'm trying to dissuade anyone in most of the US from stocking up on mining machines thinking it'll be a great heating system.
You can burn coal, yes, but once you add a chimney you're not getting 100% of the heat like you would by building a fire on your living room floor. It's an energy tradeoff vs dying of carbon monoxide poisoning. And I assume they're worse pollution-wise than a power plant.
I'm not sure what the efficiency on those is today. Article from 1980 [1] mentions 74% for a good quality stove, it's possible they've improved, but I doubt there's been as much development work on coal burning stoves as there has for gas fired furnaces.
That's pretty reasonable, but let's get really serious--You can be more efficient than that. This winter I may buy a bicycle-electric generator, an inverter, and use it to power my laptop to mine bitcoins. This heating beats a mere 98% gas heater by capturing generated heat most efficiently: Directly inside my body. The remaining kinetic and electric energy heats my house. And since I need to lose weight anyways, the cost of electricity is less than zero. Infinite efficiency!
Ah, but the total efficiency of getting and consuming calories is quite poor. Unless you eat coal, in which case it's probably other problems that are primary.
Interesting point about on-site combustion heating.
Assuming that the fossil fuel to electricity conversion is about 1/3 efficient and the heat pump CoP is 3, then on-site combustion heating is effectively as "efficient" as a heat pump.
I guess the main point is that "fossil fuel -> electricity -> grid -> heat pump" cycle seems crazily inefficient in terms of losses, especially when heating is a large portion of electricity consumption in many households.
Heat pumps would still work most of the year, even in Siberia. Different models have different operating ranges. I have 3 heat pumps that are good to -20 Celsius (-4 F), one of them even a bit colder.
At some point, yes, but if the sound is audible outside your house or the light shines out the window then it didn't turn into heat where you wanted it to.
So not 100% of that sound or light will be an exception, but some of it might.
A coworker of mine mines ETH to heat his house and I am considering that too. I am not sure how much you can save money compared to electric heating when you account for the cost of GPUs, though.
Heat pumps have better energy efficiency than electric heaters but whether that translates to cheaper costs is another thing. Heat pumps are expensive (about $3000 per pump) and they require regular maintenance (a few hundred bucks per pump every few years). They're also big, bulky and noisy and installation requires poking holes in exterior walls.
I have 3 heat pumps and 2 fire places in my house but I still do need a bunch of electric heaters when it gets properly cold. The weather around here is right about freezing this time of year and so far no electric heaters are needed.
But soon it's time to turn on the heat and it's a consideration of money, not energy efficiency.
Pretty clever, I've never thought about it like that. Basically they turned a computer into a heater that pays for itself plus some. Makes you think we should be building datacenters where the heat could be used, rather than just pumping it out.
There was an attempt to host servers in a similar setup in regular houses, not sure what happened to it (the article link doesn't seem to work, either): https://news.ycombinator.com/item?id=9334792
That was the premise behind this machine[1] (love the unintentionally humorous product rendering) but for some reason I haven't seen them popping up everywhere.
>It's not technically a measure of "efficiency", we call this the "coefficient of performance" instead.
I have to disagree. At its most general, the concept of efficiency means "desired output achieved per unit of scarce good expended" and can be applied in a variety of contexts: new users per dollar spent on marketing is a measure of your marketing efficiency, treasure value found per hour spent searching is a measure of your treasure-finding efficiency, etc.
Heat pumps are just a case where a) we use the same units for output and input even though they're different things, and b) in that unit system, the value for the output can be greater than the input. Since people can be unnecessarily suspicious of >100% efficiency, they use different terms for it.
If you expressed the efficiency as "2 joules of heat moved per joule of mains electricity consumed" you eliminate the canceling that would let you phrase the ratio as 200%.
> If you expressed the efficiency as "2 joules of heat moved per joule of mains electricity consumed" you eliminate the canceling that would let you phrase the ratio as 200%.
But that's the performance. Efficiency is usually a measure of how good the performance is.
In this case there's a practical reason not to call it efficiency, because there is a well understood maximum performance which is the Carnot limit, so for practical purposes (to avoid confusion) PumpedHeat/Power is called "coefficient of performance"and CarnotLimit/PumpedHeat is called "efficiency", and it has the usual properties that it cannot exceed 100% and it measures how well your system works in a more or less absolute sense.
Heat pump takes energy from two sources, power grid and outside environment: air, landmass or water (there are even air+land models etc). Their energy output is always less than sum of sources, so their efficency is never over 100%. But as the ambient land/air/water heat is generally ”free” then we seem to get 3-500% economy. I have never even seen the technical/physical efficiency numbers for heat pumps as consumers do not care about this.
Right, heat pumps cost 4x less, but a miner is returning profit. The only argument to be had is an ecological one. Mining is cheaper than a heat pump, gas, etc..
Of course, it would be cool if Bitcoin didn't require throwaway computation. Then we could use this computation for something that's actually valuable, like protein folding, or something, and still make use of the waste heat.
>it would be cool if Bitcoin didn't require throwaway computation
Sure, but that's a technology that does not exist. If you can figure out how to create a blockchain-like concept, where the 'proof of work' actually contributes to humanity, you have invented the next big thing.
Edit: Reading the comments below me, it looks like this is a space that is being actively explored. Exciting!
There are some coins that do something like this, primecoin's proof of work has to do with primes.
> The innovative prime proof-of-work in Primecoin not only provides security and minting to the network, but also generates a special form of prime number chains of interest to mathematical research. Thus primecoin network is energy-multiuse, compared to bitcoin.
This is of course very basic, but still, it's interesting.
You can have blockchains without decentralized consensus. It's how Certificate Transparency works.
You have the contract host record all operations and output on a ledger, and then ship that ledger/log out to monitors. This improves the auditability of the system, and enforces the backend logic as a contract. Deviations from the code are detectable by replay, and then provable. Flag the host and readjust your trust.
You can also find ways to reduce the cost of switching between traditional hosting models, such as using data-sharing networks (Dat, IPFS) to quickly "hydrate" a new host, and good ole configurable endpoints to make the switch.
It's the opposite of a trustless network; it's a trust network built on accountability. You watch hosts for bad behavior (such as the violation of the backend contract) and reconfigure trust around violations. No decentralized operation, but much less wasteful than PoW.
Folding and SETI @home existed for more than a decade. I've never understood why nobody mentions them for Proof of work. Maybe the work's not verifiable?
Yeah I believe that's the problem, the work needs to be trivially verifiable. You also need to be able to modulate the difficulty easily so that the blockrate remains constant as more people enter the "contest". And of course the work itself must somehow reference the transaction data otherwise you could use the same proof of work for two different blocks.
Seems difficult to find an actually useful mathematical problems that fits all these prerequisites.
This is actually already a thing., check out Gridcoin. Been around for a couple years but never really took off. I still think it's a really good idea, but might need better marketing.
>Curecoin puts ASICs to work at what they are good at–securing a blockchain, while it puts GPUs and CPUs to work with work items that can only be done on them–protein folding.
Do you (or anyone) have more information in how this works? Does this reward you directly for doing the folding (as opposed to validating)? How does that work? That looks like a very interesting concept, and if it's technologically sound then I'd love to see it gain popularity.
As long as your problem can be defined entirely by data from the previous block and there is a nonce you can vary you can adopt proof of work to just about anything.
The problem is that for most problems you don't want to explore the space of potential problems randomly, you want to select ones that people care about. If you allow an actor to dictate what problem to work on then you can have that actor pick a problem they have already solved and will be able to roll back tx arbitrarily once they have control.
It's not. It still uses proof of work, for now at least. They say they want to implement proof of stake. THAT will be interesting, but I want to see if they ever get there.
There was an interesting article "Nothing is cheaper than proof of work"[1] that talked about both proof-of-stake and using miners to heat your home. Basically, if you decide you are willing to pay $9 to make $10 back, then discover you are additionally saving $1 off your heating bill, you will logically be willing to pay up to $10 to make $10 back + $1 in savings.
But, the main point of the article was that "proof of stake" is basically selling liquidity (exactly like bonds) and, to get the same job done, the cost of that liquidity won't end up being any cheaper than the cost of the electricity it's replacing.
The Ethereum Foundation has published a response [2].
I suspect you're getting downvoted (though not by me) because the hashing computation doesn't directly provide that.
Other proof of work algorithms could concurrently secure the block chain and also provide direct, external benefit. There are some suggestions elsewhere in this thread. :)
Funny you're getting down-voted. This is quite literally a true statement.
Then the only point of contention is whether the benefit of this secure network existing and what it gives the world outweighs the cost of the resources consumed.
No, that's not the only one. High on the list is, surely there must be some way to accomplish the same quality of network without requiring so much throwaway work.
I mean deriving value the actual results of the computation, not the side-effect that it can serve as a competition. Hypothetically, we could derive value from both of those things.
I used to increase my PC voltages every winter at my parents wooden heated house near the arctic circle. The room temperatures could drop all the way to 16C (60F) during the night. That allowed easy overclocks on my Athlon processor from 3 to 4Ghz. I still had to wear wool socks though, because the air pushing from the central unit would cause an annoying breeze of cold air onto my toes in the always chilly room.
Hah, my place in the SF bay area often gets that cold on cool nights, no insulation, single pane windows and heat is expensive, blankets and thick socks are cheaper.
Interesting, I didn't know NYC apartments get so cold! But yeah, maybe it's just me, but I prefer sleeping in a warm (or hot) rooms.
Also, it's still actually cold outside. Coldest mornings the year I left my parents apartment were around -40C (-40F). I personally withdraw from going outside by choice after it gets below -30C (-22F). Below that point, I start having random nose bleeds of blood. People from the south may pass out.
It's very uncommon for homes to be as cold as 16C in Sweden, in fact if you rent you can demand that the temperature doesn't get lower than 18C, or 16C on the floor. It's probably because of better insulation.
The law in NYC is something to the effect that between October (I think) and March, if it's colder than X degrees outside the landlord must heat the apartment to 65F in the daytime and 55F at night.
Most of that issue is insulation (or lack thereof). Where I live (Northern Europe) it's been 0C-10C/32F-50F outside for the past couple of weeks, yet I still haven't turned the heating on. Inside temperatures have been a steady 21C/70F.
My apartment was newly-built last year, and rated B for energy efficiency. We have a heat-recovery ventilation system which I assume helps a lot, as we don't need to leave the windows open for fresh air, and heat from showering and the dryer is used to heat the inside air, rather than just being vented outside.
I remember in college that we found it was cheaper just to use our computers to heat our crappy rented pad than to pay for heat and the electricity for our PCs.
So - I actually physically run 2 mining rigs - each with 6 Nvidia GTX 1070s - pulling 1020 Watts each. They each heat a room fairly nicely and I make around maybe $300 in profit from just selling coins per month. This is AFTER I subtract my electricity costs which for these 2 rigs is about $160 per month. I have not quantified the heating benefits but I'd estimate it at maybe $20 a month. I built them back in March/April for $6k total, and I've made twice that so far - paying for itself already and extra $$$ - plus I still have the hardware going or I can resell it at a profit even thanks to the shortage of high end video cards due to all these crypto miners like myself.
In the summer, I had the rigs in 1 room air cooled with just the windows open - no problems with 90 degree F days at all. Fun project, and so far quite successful.
But if you'd bought $6k worth of ethereum outright in march it would be worth $90k now. That's why I haven't started mining, it only seems profitable in a sideways market.
I could always resell the video cards I bought at near cost. Buying speculative coins is another matter entirely. Further, I'm not mining ETH - if I took your advice on the coin I am mining I would have a net LOSS not gain. You should take another look at the economics of mining, I think you have a few aspects misunderstood.
I could always resell the video cards I bought at near cost
You sure about that? There are so many miners these days, if the mining market tanks, the used graphics card market will be flooded and prices will likely plummet.
I used to mine hardcore back in the good ole days in 2013, first altcoin boom. What are you hobby miners mining these days? I know whattomine.com has charts, but what are some of the favorites these days for GPU and CPU coins that are actually profitable? Monero? Back in my day it was Dogecoin lol
But that’s also survivorship bias. If you buy a rig, you can always have the flexibility of mining whatever’s profitable, and even potentially not mining when there’s no profit to be made.
12 GTX 1070 (500$ avg) + the rest makes for a ~8000$ setup right ? so basically 2 years of 300$ profit from mining to get that money back. Will the profitable mining last longer than that ?
I made it for $6k, and the glory days of mining are over, but for some time I was making $100 a day mining. That is no longer the case. I actually bought 14 1070s, but I'm using one now for gaming haha - I figured I earned it. But I bought my cards when they were $300 not $500 back at the start of the year.
I make no claims about what happens in the future though.
There is a startup which installs their servers in private homes and other buildings where they are used for heating. They essentially connect the market of server computing and heat.
This reminds me of a story; a buddy of mine during my college days lived in a solitary townhome adjoining a million-watt microwave emitter room. The thing produced enough waste heat to warm the apartment in the dead of winter, even with all the doors and windows open, plus the adjoining warehouse which was several times the volume of the dwelling.
I was envious at the time, although now I wonder about the health implications of the setup, and whether it was strictly speaking legal. The guy is alive and well today, fortunately.
It was probably fine; microwaves are fairly easily blocked by even a very thin metal screening. The main health issue with microwaves is the heating caused by it.
The eyes are usually what is damaged first because they tick all 3 boxes of:
* Easily damaged by heat
* High water content
* No good way of shedding heat
Testicles hit the last two, and the degree to which the first is true is debated.
> It was probably fine; microwaves are fairly easily blocked by even a very thin metal screening.
That makes sense in retrospect, given we humans routinely stand with our noses pressed against thinly-screened microwave oven windows waiting for food to heat, apparently without health effects!
Your description makes it sound like the most bizarre mixed use building. Townhouse/industrial equipment shed/warehouse is quite the combo. I can't imagine how they classified that for tax purposes or what the land was zoned for.
Assuming the actual emitters were high above his head and fairly directional then it probably wasn't a health nightmare, the heat would have mostly been waste heat from transformers, but it's probably not ideal.
What did he do in the summer? I guess it's school so he wouldn't be there for the hottest part, but there is still part of August and September as well as May and possibly June.
Append a giant concrete warehouse to the side and move the tower into the back yard and, yes.
> Your description makes it sound like the most bizarre mixed use building.
Agreed on all points :) I should ask next time I see him about the zoning. I speculate it was some kind of employee/maintenance quarters. His job was to go into the transformer room once per day and make sure the green light was still blinking. (Kidding, not sure what the arrangement was :)
> Assuming the actual emitters were high above his head and fairly directional...
The tower was quite tall, and I was told perfectly safe to climb for most of the height, then suddenly fatal! I understood it to be more or less horizontally omnidirectional. I'm definitely not an expert in the tech. Also I think he could redirect the heat to the outside during warmer months.
> I was told perfectly safe to climb for most of the height, then suddenly fatal!
I used to know a couple of radio engineers, and this sounds about right. The owners of the radio stations would pay them extra if they did tower maintenance without shutting the station down. They all had weird swirly RF burn scars on their forearms.
> His job was to go into the transformer room once per day and make sure the green light was still blinking. (Kidding, not sure what the arrangement was :)
Ah, but did he have a dog whose job was to bite him if he tried to touch any buttons?
I think that's essentially what they did. I got to go into the transmitter room and see the 1' diameter copper pipe which I was told carried the radiation out and up to the tower, and which not to approach too closely.
As a kid I always tried to convince my mom to let me play computer games to heat our house during winter. She didn't believe me. But it was probably not the efficiency of the heating that stopped her from letting me game all day, though
Did you tell her that you move you body when you're really engaged and generate additional heat by burning your calories? That might swing the balance to your favor.
Body heat is actually 2-3 orders of magnitude from being cost-effective, unless there's a food source much cheaper than typical lard. Edibility adds quite a premium to calories.
There are many other cases of people heating homes with mining rigs, especially in places that has a year-round cold climate. Electricity savings in cooling the rigs as well as heating the home + the profit from mining, makes the whole thing worth it...
... IF you can tolerate the noise that the rigs generate
There are passive cooling rigs for high performance computers (LinusTechTips video). If this is a valid market, one could make prebuilt silent rigs as radiators.
Towards the end of the video, the owner makes a point that this democratic distribution of mining power instead of large miners controlling the network is a good idea. Curious to hear more about this -- does anyone know good writing re: Democratic / distributed benefits of mining?
The bitcoin protocol is somewhat at odds with the desired goal of distributed mining. Only 1 entity gets the reward of mining a block, and only 144 blocks are mined per day. You are encouraged to join the strongest pools to have the highest chance of making money. According to blockchain.info, the top 6 mining pools control 75% of the market.
Statistically the reward is the same whether you're solo mining or in a pool. Actually less in a pool because they take a cut of the profits. The benefit of being in a pool is that you get the (smaller, since they're shared) rewards more frequently instead of waiting possibly years until you solo mine a block.
But there's really no reason why you have to be in the biggest pool, if you're in a pool with a 1/144 share of the hashrate you'll already be getting daily income, why do you need it to be more frequent?
Note: I think solo miners also get a bit more average latency when receiving bloks, I'm not sure if that offsets the cut the pools take enough to make the expected value of being in a pool higher or not.
Note that from this perspective, a mining pool is effectively a single miner - so a Serbian miner might not actually help, depending which pool it's attached to.
> Note that from this perspective, a mining pool is effectively a single miner....
That isn't really true. A pool operator who tried to abuse their position to carry out a 51% attack against the interests of the pool members wouldn't retain their majority for long. Individual miners would switch to a different pool, or simply shut down their hardware. Conscientious pool participants could also scan the blocks they're asked to work on to detect such attempts before they are carried out, though I don't think any of the standard pool software does that yet. To sustain a 51% attack for a meaningful length of time you would have to own the hardware, not just manage it for others.
This technique could pair nicely with another problem we have- storing solar power.
In theory, you could build a system which uses solar power to mine bitcoin during the day, and store that heat somewhere (like an underground tank of water) until you need to warm your house at night.
Of course its not a perfect solution because the coldest places are obviously also the places which get the least sunlight.
> and store that heat somewhere (like an underground tank of water) until you need to warm your house at night.
Just install a solar water system if you want to heat your house, it'll be more efficient at generating heat than solar panels.
> This technique could pair nicely with another problem we have- storing solar power.
Actually the ability to quickly turn on and off mining operations has other more practical uses for excess energy. Right now there are situations during the day where utilities have to idle base-load plants because renewable energies produce enough to satisfy grid demand, or they even have to pay another utility to take their excess power in case they can't quickly throttle their base load (e.g. with Nuclear).
Having a large crypto currency mining pool the utility could use during these times would enable them to use the excess energy and generate a profit from it, instead of paying another utility to take their excess energy.
Hm, I've politely ignored bitcoin for a long time, but I live in a cold climate and do make some use of electrical heating. Is there a good user-friendly miner setup I could use to supplement a space-heater? I assume a low-power linux box running some bitcoin-oriented package with a bunch of asics shoved into the USB ports?
Even using a time-of-use plan like E6 (which they don't allow any more), off-peak is $0.16728/kWh in the summer, and peak summer is $0.35933/kWh. The ETOU-A and ETOU-B plans they're pushing people toward bottom out at $0.17279/kWh winter off-peak, up to $0.36335/kWh. Again, all baseline figures. Electric Vehicles and over-baseline (read: mining) will punish you to death.
In those rate plans, PG&E apparently spends more on transmission alone than Siberian customers entirely pay for service. Also, good luck explaining the plan structure to someone; it took me 20 minutes of edits just to put this comment together coherently.
That's why SolarCity PPA is so popular - the whole idea is that they shift most of your electricity consumption back into the 1s tier. Probably still not enough for mining though unless you can make a deal where you get way over regular capacity in solar for the same price.
Siberia has huge rivers which are very suitable for building hydroelectric because the accompanying flooding happens in mostly uninhabited areas and thus has relatively little costs.
In my experience, most leases that include utility costs will often contain language allowing them to charge you for what they consider excessive or unreasonable usage.
There's nothing new about this. I remember a story from the days of mainframes, where a data center had to be redesigned - they had built it without heating, expecting the mainframe to provide all the heat the building needed in the winter. By the time they got around to buying a machine to put in it the technology had progressed and it was much more power efficient than they had planned.
Those days aren't over. Mainframes are still in serious use, at least in banks, and (I didn't know this until going to find the following link) apparently also have Blockchain support: https://www.ibm.com/us-en/marketplace/z14
ASPLOS 2008 conference on Wild and Crazy Ideas already suggest the more general version of this: "Computation: A Byproduct of Home Water Heaters" http://cseweb.ucsd.edu/~swanson/WACI-VI/
Earning $430 for month by bitcoin mining. What about the electricity costs?
If you factor in electricity costs and the money they spent on infrastructure(GPU's and high-end machines), it is a dumb idea. They could have spent $30-50 on heating equipment instead.
Electricity costs are probably subsidized (since otherwise in places where it gets super-cold it wouldn't look nice if people froze to death for inability to pay their power bill).
Electricity costs in Irkutsk (place where the place that the video is describing is located) are 0.7 RUR per 1 kWh for rural areas. That's about 1 cent per kWh. For comparison, California's lowest tariff in about 20c per kWh and if you are consuming a lot (which you'd do if you mine) you'll get over 35c/kWh. If you qualify for low income subsidy, though, you might get it for significantly cheaper :)
Is there a way into mining that is scalable? I mean some way where you don't have to shell out $8k up front and can build-as-you-save? There are lots of ways to save money on heating for people with $8k to spare
Same as any other rig for anything. Higher load usually doesn't make much difference in lifetime.
The problem with components is that they can break one day after you bought it, or last 50 years.
For example, in CERN they bought 100,000 disk drives for their huge amount of data. Lets say disk lifetime is 1M hours which is 114 years, so sounds pretty good. But that means, on average, every 10 hours one of their drives will fail.
O once had a little freelance gig for a company that did offer computing resources to one part of their customers and heat for houses and water to the other with just that concept. cloud & heat
Check out electricity prices in SF. Especially the top tier. You probably would be better off burning small-denomination bills for heating. Unless of course you tricked your landlord into covering electricity costs, in which case mine away until police comes to evict you :)
I can attest, I have a 9 card mining setup which pulls around 330w per card 24/7. In the summer when it was 70-80 degrees outside it was enough to keep a large basement in the 90s, plus the rest of the house uncomfortably warm. I had to purchase an AC to keep the place livable.
Now that winter is coming around I look forward to the extra heating.
There seems to be discussion in these comments about mining for the purpose of heating a home, but the heating is a bonus. You're making money AND saving on energy, that's the important part.
Plus, in my opinion, a constant heat source like a mining rig is preferable to an older cycling central system which can be slightly disturbing as it loudly cycles on and off throughout the day...
Electric heaters are 100% efficient at turning electrical energy into heat. All the energy they take in, that's where it goes.
Bitcoin mining (and computers in general) are basically in the same boat. There's some ancillary energy use for blowing air around or shining lights or making sound waves, but for the most part the electricity turns into heat and does some math along the way.
Sounds great! Perfectly efficient! But the important caveat is that heat pumps are (effectively) more than 100% efficient. They can take 100 watt of electricity to transfer 300 watts of heat from outside to inside.
It's not technically a measure of "efficiency", we call this the "coefficient of performance" instead. Or we call it the "seasonal energy efficiency ratio", which is the same thing except multiplied by some constant factor because somebody wanted a scale that went up to higher numbers. Or for heating, the heating season performance factor, which I forget the specifics of, but is geared toward how well it works pumping heat in from outdoors when your heat source (outdoors) is very cold.
In short, it's just as efficient as any other electric heater, but heat pumps are 3-4x better than that.