Doing computations hardly stores any energy anywhere (the energy stored in, say, magnetization in a magnetic disk is pretty much negligible), so almost 100% of electricity is turned into heat.

The problem with this ~100% efficiency is that, if your goal is heating, you can move way more than 100% heat with 100% electrical energy if you use, say, a heat pump.

Heat pumps are not viable in a lot of cases, in an urban setting, it's probably the vast majority. So for the many cases were the alternative for heating would be burning fossil fuels, using electricity, ideally produced via renewable energy/nuclear, could be a superior alternative.

Wait, why heat pumps are not viable in an urban setting?

My urban apartment has a heat pump and it's nothing special. I even pay the heat bill and they still gave me a heat pump. I have no idea why someone would say they aren't viable. Not only are they viable, they're typical for new construction in warmer climates.

For many reasons.

First, transporting heat as opposed to electricity is very wasteful, so you only want to transport it in very short distances.

Second, typically even small single family home requires quite large volume to store the heat effectively for many months. It isn't that complex only because in a single family home setting you already have a bunch of uncontested land available so you can use the volume that is relatively flat and not too deep.

Building this on a scale of a city would be insurmountable challenge. You would have to dig deeper than the buildings are high and any kind of works like that are difficult in urban areas.

A heat pump is just a backwards air conditioner. Not whatever you are thinking it is.

I think you are thinking of something very different than what "heat pumps" actually are? They don't involve storing heat, they're just an inverse fridge.

I think guys you are all wrong.

https://en.wikipedia.org/wiki/Heat_pump

The first sentence:

"A heat pump is a device that transfers heat energy from a source of heat to what is called a thermal reservoir."

So yes, it involves energy storage.

The way this works is you store heat in the summer (warm up a lot of rock or ground underneath your house) and recover that energy in the winter by pumping a liquid through warm rock back to your house and use it as a heat source.

The air outside your house counts as a thermal reservoir as well. That it doesn't store anything to the next season is annoying and makes it not work well in cold temperatures.

I think wikipedia is wrong or confusing here. A heat pump requires a thermal resevoir (something that doesn't change temperature much when you move heat to or from it) of some sort, usually the atmosphere, or in ground source heat pumps pipes running through the ground, but it can move heat in either direction.

> While air conditioners and freezers are familiar examples of heat pumps, the term "heat pump" is more general and applies to many heating, ventilating, and air conditioning (HVAC) devices used for space heating or space cooling.

Basically what I would understand from the term heat pump in ordinary conversation would be an air conditioner intended for use in a heating dominated climate rather than a cooling dominated one, but there might be some regional differences in usage.

Thermal reservoir is a technical term in thermodynamics and as others have pointed out it doesn’t mean what you think it means. Anything remotely resembling a (reversed) Carnot cycle would involve thermal reservoirs. You can read its own Wikipedia page.

Given that the neighbouring house has one on their wall, I think I know what they are.

What a few of these "lot of cases" where they are not viable?

Anytime it is "cold" air source heat pumps don't work. Cold varies a bit, somewhere between -5C and - 25C depending on design factors. Even in the best case as you get closer and closer to the minimum temperature the worse they work (IE when you need them most!), and once you hit the cut-off you better have a backup source off heat.

You can use geothermo (ground) to work around this. I'd recommend it, but the one time install costs mean it is questionable if it is cost effective.

OK, so that's one, very well understood case. Yes, a heat pump will not work all the time. I have one and it stops working around 20F and the furnace takes over. So what? I live at the US/Canada border and my furnace runs maybe ~2 months during the year. This is still tremendous savings.

Big savings, but is it big enough to be worth the extra expense of a heat pump vs used using the furnace year round. The times when the heat pump works are the times when you least need it, since other activities of life are adding heat to the house too.

When I said Data Center can produce electricity I did not mean that it can produce as much or more than it takes in. I thought this obvious enough that it did not have to be said.

But alas, it has to be said.

No, you can't create perpetuum mobile ie. build a machine that given a supply of energy produces as much or even more energy.

https://en.wikipedia.org/wiki/Perpetuum_mobile

It is still not viable.

Assume your datacenter runs at 50C (122F) and the temperature outside is -40C (-40F), using the datacenter heat to generate electricity has theoretical maximum efficiency of 27.85%. If your datacenter is at 23C (73F) and room temp is 0C (32F), then the theoretical maximum is 7.77%. Note the word 'theoretical maximum', in reality probably at least 10 times worse than that.

You assume DCs are cooled by the whole volume of air inside when in fact you can put heat pipes on components and get closer to 70-90 degrees heat source. Even years ago when I worked in datacenters the air was circulated from outside to the server rack and then back outside, never crossing from the rack into server room.

Newer generations of CPUs are going to be exchange less frequently but will produce more energy as they offer more dense computing. This makes case for investing more in the server hardware.

Even then understand, that 7% of a huge amount of energy is still huge amount of energy.

Edit: Apparently "Facebook datacenter in Denmark is supplying around 10.000 homes with heating" -- source, another poster.

So... you need to rethink your expertise on defining what is and what is not viable.

You can supply heat no problem. What I was talking about was recovering energy from heat, which is physically limited by Carnot engine. The percentage was Carnot engine efficiency. Note that it is impossible to create Carnot engine in real life, so actual efficiency is much, much, much lower.

No, essentially 100% of energy used for computation will turn into 'waste' heat.

If you calculate 1000 digits of pi, those digits will not embody any energy.

Just turning chaos into order takes thermodynamic work, but I was wildly off with my percentages. Thinking in terms of an idealized computer that doesn't create heat, but didn't realize how far we were from that.