I've always found it interesting that electric radiators are 100% efficient but if you treat a computer as a radiator then you get the same 100% efficiency but extra computations done on top.
Was that a joke? Both points are wrong. Electric radiators aren't 100% efficient, can't be. e.g., some of the transformed energy will be light (the metal gets red when it starts to get hot).
And of course if you use a computer then X% will go into the computation. The reason why it might be cost efficient for you is because the server guy thinks he has the same costs for electricity no matter if the server is in the data center or in your home, and therefore pay your bills for the server heater.
On your second point, no, a 1000W heater will emit the same amount of heat (neglecting visible light losses, which, as other commenters point out, will be converted into heat later on) as a 1000W computer. This is the first law of thermodynamics, and the fact that the computer is doing computation makes no difference.
Perhaps, but the one thing I find more annoying than people spreading misinformation is when people incorrectly accuse someone else of being incorrect.
However, I will admit that my comment did not add anything useful to the discussion.
Part of the energy can get transformed into light (mostly infrared) or sound, but unless it goes out by the window, all that energy will bounce until it's absorbed by a the walls and furniture. It's very easy to get 100% electricity to heat conversion if you have not windows.
You get also 100% conversion with a normal lamp or a sound system (unless they are very close to a window). You get all the heating, and the light or music for free.
I investigated further into that. The confusion arrised from unclear wording in my memory, I guess. You are right electricity to heat is 100%. The inefficiency of electric heat radiators is in the electricity production being inefficient compared to burning the same resources for heat directly.
Physics must have a word for how energy in forms such as electricity can do work, but once transformed into thermal energy it can no longer do work (a temperature difference is required to perform work, and then only at carnot efficiency — a temperature alone cannot perform work).
I asked on reddit once, and someone suggested "exergy" but I think that word is explicitly concerned with heat, so couldn't be used to say things like "1000w of electricty provides more exergy than 1000w of gas", or "An electric space heater destroys the exergy of the electricity it consumes, while a computer uses it for computation".
"Energy is conserved, but it can change forms. If you have energy in a low-entropy form, you can do useful work with it. You can lift something up, drive a car, or fly an airplane. If you convert that energy into a high-entropy form, it becomes useless.
A low-entropy concentration of energy is called fuel. We have fossil fuels sitting in the ground with energy in them in a concentrated form. We can extract the energy because the entropy of the fuel is low. Once we burn the fuel, we cannot go back. You can heat a room in your house by burning wood, but you cannot cool off a room in your house by unburning fuel and turning it into wood."
We do have some startups in Germany, that do that kind of thing already. So their story seems likely to be PR BS to me.
I was asking myself the question, if I would trust my data to be stored around the country in different private homes. And as a user of a heating solution, I would ask myself, how secure their business model is. If i switch from gas to server based heating, I would have to be sure, that 5 years from now, I would not have to shoulder the costs of switching back (one provider lets you have free heating for the costs of switching from gas/oil to server based).
So as interesting, as I find these ideas, as a homeowner, I would not want to have to switch heating during midwinter because the cloud provider goes down the drain.
Why would you replace your old heating system? I would think of that as an additional heat source that enables you to save some resources from your general heating system.
Think of it as the solar water heater on a house's roof. It doesn't always works but when it works it helps you save some money on other options.
Well the ones I found in Germany need you to change the system, as they do not provide heat directly in the rooms, as in the OP, but have a central system (server rack) that heats the whole house/flat and docks onto the existing piping.
Maybe the key to this market is to view it partially as charitable gift to your favorite research project, SETI, protein folding etc. This data is not particularly sensitive and can run with little network activity.
So I looked into doing this on a personal level a while back and did some back of the envelope calculations and it was hard to square the economics of it.
Two factors made it really difficult:
1. All computation is done in DC. Even if you assume all other prices fall to 0, the AC->DC conversion cost on a per watt basis makes it hard to compete with cheap heaters which can operate on pure AC. Just power supplies capable of powering 1000W of computation costs a lot more than a 1000W heater, even used. You need to get the economics exactly right to make this scheme profitable and the factors are so sensitive to shifts that it's hard to plan for the capex.
2. There's a noise/convenience/space tradeoff that's hard to square. Ideally, you want the heat to be close to the users but the noise to be far away. It's hard to make electronics perform well above 100C as opposed to resistive elements that work fine up to 1000C. That means either moving air quickly (noise) or moving it over a large area (space). Neither of these are ideal.
It's possible for these two factors to be overcome but they're sizable and makes me uncertain whether it'll ever be possible to scale this as a viable business vs a cool idea that gets trotted out in trend blogs once every few months.
The whole trick about any kind of 'co-generation' is that you get something you need (computation) and something that you can use that would otherwise go to waste (heat).
The only time that scheme does not work is if you don't need the heat.
Right, but the profit from co-generation from 100W of computation has to be at least greater than (cost of generating 100W of DC) - (cost of providing 100W of pure heat).
I was looking at it in the context of bitcoin mining at the time but it's remarkably hard to find computational tasks that both generate good $/FLOP returns and are stable and scalable which is what a scheme like this needs.
I'm a bit confused, don't the losses of an AC->DC converter go into heat as well? Why is that less efficient than a cheap AC heater? Is the energy lost in some other way or is the heat generated by the converter not usable?
No, the profit of co-generation from 100W of computation has to be greater than (the productivity loss from the geographic dispersion of the computational resources minus the cost of actively cooling the computers).
The electricity is free because it would already have been spent on computation (plus some more on cooling that computation, which is no longer necessary).
What about the energy conservation. Of course there are energy leaks in AC->DC conversion if you don't take into account the heat produced during the conversion. If your purpose is heat, the efficiency is 100%.
A radiator behaves like a very big passive cooler for your data center. You want it to be less than 60°C to avoid getting burned. That is perfect for CPU. Fans are definitely far from mandatory.
Depends on outside temperature, if it's cold enough heat pumps become less efficient than simple resistance heating. The transition point where they become useless is generally around 0F which is not all that cold. Granted if you live in a warm area then it's generally not an issue.
> To prevent the server stacks overheating, tech companies spend vast sums on cooling technology - more than a third of a data centre's hefty energy bill may go on air conditioning.
Data center cooling isn't necessary. I saw a presentation by Christian Belady, the general manager of Data Center Services for Microsoft's Global Foundation Services (GFS) group, where he spoke about how they don't cool their data centers anymore. He tested the theory by putting a bunch of racks outside in a tent to see that they wouldn't overheat.
Sounds like a good idea, although it appears to still use convection heating, which is the least sustainable form of heating. Low Tech magazine has some relevant articles related to the topic[0][1].
Anyway, back when I still used desktop computers I always made sure they were placed in such a way that they doubled as a heater for my feet. Now that I think of it, it probably would have combined quite nicely with a kotatsu-style set-up[2].
I'd rather see this kind of thing heating water, which could then be used either for space heating or directly as hot water (in both cases, as input to a conventional boiler, rather than being expected to provide all the heat itself). In places with large seasonal temperature variations, one typically still wants hot water for washing, etc. even when space heating would be undesirable.
To combat seasonal demand, last-mile, and access control issues, I would actually want to see this set up either as a district heating scheme* (e.g. in the large buildings built for mechanical telephone exchanges); or to heat water for municipal or club swimming pools.
Some ex soviet countries have city central heating systems. I am from Vilnius, Lithuania, we had that in our old flat. Maybe it might be a good idea to create data centers around there?
The only unsolved (as in I don't have an idea) problem is, what do we do with all the heated water in summer?
That's the very reason for heating water instead of space, hot water is wanted year-round.
It's used for washing both people and things; Swimming pools are normally heated, even just a little, above the cold water mains temperature. Hammams need hot water, even in the height of summer.
Hot water can be stored, out of the way, in an insulated tank, for when it is required.
Instead of providing fully-hot water to a few very nearby households, it can be used as a warm-water input to domestic boilers for a larger number of households over a larger area.
Even in a self-contained domestic system, it would be better to pre-heat water going into the boiler, than to just be a radiator that heats the outside in summer.
"user needs heat but the internet is down and the radiator has nothing to work on, it starts performing dummy equations" - even bitcoin mining is greener than this
It doesn't go in to detail about how it's free to use. Presumably as well as a fibre optic cable it needs to be plugged into the grid, so do you need to invoice them for power costs at the end of each month?
This seems like an interesting, but inefficient solution. (What do you do during the summer when heat isn't needed? How much more do these units cost over a regular computer at a data center? Who is paying for the electricity and bandwidth for these machines?)
Is there some sort of complementary process that could use the heat generated by the data center at the site itself? (energy production, manufacturing, etc)
During summer it will dump heat outside through the wall. That's why the device requires to be installed on an external wall. That's actually quite smart, if they can implement the actual mechanism inexpensively and reliably.
Yeah that's much more interesting. There's a limited need to do these things locally. Not just data centres btw, factories, too, actually mainly.
But that requires heat pipes throughout cities, basically next to your electricity, internet, water & sewage, you now want additional heat pipes running through every home. That requires huge infrastructure investments, which makes sense for new real estate, but is probably not worth it for existing homes.
Lower hanging fruit is actually other factories. The infrastructure to connect these two is much more trivial in scale, as well as in execution (laying a pipe in sandy industrial land isn't as big a deal as under paved residential areas). For non-factory buildings like swimming pools, this works great, too. So in the Netherlands for example there's one milk factory which heats and cleanses swimming pool water. (the pool was built with this in mind though, not retrofitted). And one large elderly home is heated by waste heat from a sewage cleaning facility. Another entire industrial area of factories is supplied with the heat from a waste burning facility. All of this is much easier than tackling homes, for now.
50 comments
[ 44.7 ms ] story [ 299 ms ] threadAnd of course if you use a computer then X% will go into the computation. The reason why it might be cost efficient for you is because the server guy thinks he has the same costs for electricity no matter if the server is in the data center or in your home, and therefore pay your bills for the server heater.
However, I will admit that my comment did not add anything useful to the discussion.
You get also 100% conversion with a normal lamp or a sound system (unless they are very close to a window). You get all the heating, and the light or music for free.
Physics must have a word for how energy in forms such as electricity can do work, but once transformed into thermal energy it can no longer do work (a temperature difference is required to perform work, and then only at carnot efficiency — a temperature alone cannot perform work).
I asked on reddit once, and someone suggested "exergy" but I think that word is explicitly concerned with heat, so couldn't be used to say things like "1000w of electricty provides more exergy than 1000w of gas", or "An electric space heater destroys the exergy of the electricity it consumes, while a computer uses it for computation".
So is there a word?
"Energy is conserved, but it can change forms. If you have energy in a low-entropy form, you can do useful work with it. You can lift something up, drive a car, or fly an airplane. If you convert that energy into a high-entropy form, it becomes useless.
A low-entropy concentration of energy is called fuel. We have fossil fuels sitting in the ground with energy in them in a concentrated form. We can extract the energy because the entropy of the fuel is low. Once we burn the fuel, we cannot go back. You can heat a room in your house by burning wood, but you cannot cool off a room in your house by unburning fuel and turning it into wood."
I was asking myself the question, if I would trust my data to be stored around the country in different private homes. And as a user of a heating solution, I would ask myself, how secure their business model is. If i switch from gas to server based heating, I would have to be sure, that 5 years from now, I would not have to shoulder the costs of switching back (one provider lets you have free heating for the costs of switching from gas/oil to server based).
So as interesting, as I find these ideas, as a homeowner, I would not want to have to switch heating during midwinter because the cloud provider goes down the drain.
Think of it as the solar water heater on a house's roof. It doesn't always works but when it works it helps you save some money on other options.
Two factors made it really difficult:
1. All computation is done in DC. Even if you assume all other prices fall to 0, the AC->DC conversion cost on a per watt basis makes it hard to compete with cheap heaters which can operate on pure AC. Just power supplies capable of powering 1000W of computation costs a lot more than a 1000W heater, even used. You need to get the economics exactly right to make this scheme profitable and the factors are so sensitive to shifts that it's hard to plan for the capex.
2. There's a noise/convenience/space tradeoff that's hard to square. Ideally, you want the heat to be close to the users but the noise to be far away. It's hard to make electronics perform well above 100C as opposed to resistive elements that work fine up to 1000C. That means either moving air quickly (noise) or moving it over a large area (space). Neither of these are ideal.
It's possible for these two factors to be overcome but they're sizable and makes me uncertain whether it'll ever be possible to scale this as a viable business vs a cool idea that gets trotted out in trend blogs once every few months.
The only time that scheme does not work is if you don't need the heat.
I was looking at it in the context of bitcoin mining at the time but it's remarkably hard to find computational tasks that both generate good $/FLOP returns and are stable and scalable which is what a scheme like this needs.
The electricity is free because it would already have been spent on computation (plus some more on cooling that computation, which is no longer necessary).
A radiator behaves like a very big passive cooler for your data center. You want it to be less than 60°C to avoid getting burned. That is perfect for CPU. Fans are definitely far from mandatory.
Data center cooling isn't necessary. I saw a presentation by Christian Belady, the general manager of Data Center Services for Microsoft's Global Foundation Services (GFS) group, where he spoke about how they don't cool their data centers anymore. He tested the theory by putting a bunch of racks outside in a tent to see that they wouldn't overheat.
At first, I thought it would be simple, but soon realized there are many, many factors involved
Anyway, back when I still used desktop computers I always made sure they were placed in such a way that they doubled as a heater for my feet. Now that I think of it, it probably would have combined quite nicely with a kotatsu-style set-up[2].
[0] http://www.lowtechmagazine.com/2015/02/heating-people-not-sp...
[1] http://www.lowtechmagazine.com/2015/03/local-heating.html
[2] http://en.wikipedia.org/wiki/Kotatsu
To combat seasonal demand, last-mile, and access control issues, I would actually want to see this set up either as a district heating scheme* (e.g. in the large buildings built for mechanical telephone exchanges); or to heat water for municipal or club swimming pools.
* https://en.wikipedia.org/wiki/District_heating
The only unsolved (as in I don't have an idea) problem is, what do we do with all the heated water in summer?
It's used for washing both people and things; Swimming pools are normally heated, even just a little, above the cold water mains temperature. Hammams need hot water, even in the height of summer.
Hot water can be stored, out of the way, in an insulated tank, for when it is required.
Instead of providing fully-hot water to a few very nearby households, it can be used as a warm-water input to domestic boilers for a larger number of households over a larger area.
Even in a self-contained domestic system, it would be better to pre-heat water going into the boiler, than to just be a radiator that heats the outside in summer.
Seems like the model is something that will become more familiar in the future.
Is there some sort of complementary process that could use the heat generated by the data center at the site itself? (energy production, manufacturing, etc)
But that requires heat pipes throughout cities, basically next to your electricity, internet, water & sewage, you now want additional heat pipes running through every home. That requires huge infrastructure investments, which makes sense for new real estate, but is probably not worth it for existing homes.
Lower hanging fruit is actually other factories. The infrastructure to connect these two is much more trivial in scale, as well as in execution (laying a pipe in sandy industrial land isn't as big a deal as under paved residential areas). For non-factory buildings like swimming pools, this works great, too. So in the Netherlands for example there's one milk factory which heats and cleanses swimming pool water. (the pool was built with this in mind though, not retrofitted). And one large elderly home is heated by waste heat from a sewage cleaning facility. Another entire industrial area of factories is supplied with the heat from a waste burning facility. All of this is much easier than tackling homes, for now.