8
$\begingroup$

I have been wondering if it would be a good idea to exploit the natural coldness of space to create the conditions needed for quantum computation at a large scale.

The average temperature of space is about 2.7 kelvin as far as I know, which would mean that it would require a lot less powerful cooling systems to get to the temperature needed for quantum computations which is about much closer to absolute zero.

I am aware that the computer would require shielding against cosmic rays, against the radiation coming from the sun and space and finally a considerable set of radiators to dissipate the heat efficiently, however assuming that the civilization to construct the orbital computers had such means of shielding (in the form of powerful magnetic shields) and space engineering, would it make sense for them to go through this endeavor for quantum supremacy?

And most importantly, would the smaller temperature gap make a real difference?

$\endgroup$
20
  • 15
    $\begingroup$ Do note that being in a cold medium isn't the only thing. You need to be able to transfer the heat. If you have barely any medium at all to that you want to take up the access heat it'll go badl as if you have a computer in a tiny box with cold air. Soon enough the air will be saturated with heat. $\endgroup$
    – Trioxidane
    Jul 3 at 9:18
  • 7
    $\begingroup$ It's almost certainly easier to keep something cool on Earth, where you can drive up with trucks full of refrigerant whenever you want, than in space where your supply of consumables is sharply limited. $\endgroup$
    – Cadence
    Jul 3 at 9:21
  • 1
    $\begingroup$ I was more thinking that the moons might be more practical. Quite a few to chose from. $\endgroup$ Jul 3 at 9:39
  • 4
    $\begingroup$ Putting in space a shield good enough to dampen gamma rays (magnetic shields won't work) might cost more that helium cooling on Earth. 30 to 60 meters underground, not too deep to avoid geothermal heat, but enough to provide extra shielding would be more effective. $\endgroup$
    – FluidCode
    Jul 3 at 14:33
  • 7
    $\begingroup$ cooling isn't about having a cold medium. it's about how fast you can transfer heat to it. Compare: 15 degrees air vs 15 degrees water, metal vs plastic... some feel cooler, because they are capable of transferring heat faster. Vacuum is not very good at transferring heat (no conduction, no convection) $\endgroup$
    – njzk2
    Jul 3 at 17:33

8 Answers 8

24
$\begingroup$

2.7 K is the background temperature of space. In Earth orbit, you've got a huge pile of fusing plasma 1 AU away, and a big warm planet filling nearly half the sky. The equilibrium temperature of an inert object in LEO is about a hundred times that, and that of an object that's actively generating heat is even higher.

Take JWST as an example: it's an infrared telescope that needs to be kept very cold. Passive cooling is part of the solution, but requires it to be located at the Earth-sun L1 point, 5 light seconds away, where a shade could protect it from the sun without Earth heating it from behind. This approach wouldn't work in low Earth orbit, and even with an elaborate multi-layer sunshade doesn't get things as cold as they want. To get the coldest parts to operating temperature, JWST still needs active cooling, which is actually easier on Earth's surface where there's more power available, less stringent mass budgets, and air and water to remove heat.

In the end, the sun-facing side of JWST is currently 318 K, and even the coldest non-actively-cooled part is 37 K, over 13 times the 2.7 K background temperature.

$\endgroup$
5
  • 3
    $\begingroup$ "As hot as a Rotisserie Chicken" I believe is the term for everything in orbit. That's their temp. $\endgroup$
    – J.Hirsch
    Jul 3 at 18:25
  • 3
    $\begingroup$ These issues were solved for the JWST, which is cooled by being exposed to space in the way OP suggests. (That's not to say it's practical for quantum computing necessarily, but it's clearly not impossible.) $\endgroup$
    – N. Virgo
    Jul 3 at 22:09
  • 11
    $\begingroup$ @N.Virgo part of the solution for JWST was to put it way off at the Earth-sun L1 point, 5 light seconds away, so a shade could protect it from the sun without Earth heating it from behind. JWST's approach wouldn't work in low Earth orbit. Another part of the solution for JWST is active cooling, and that works better on Earth's surface, where there's more power available and air and water to remove heat. Also note that the sun-facing side of JWST is currently 318 K, and even the coldest non-actively-cooled part is 37 K, over 13 times the 2.7 K background temperature. $\endgroup$ Jul 4 at 2:20
  • 1
    $\begingroup$ I don't disagree - your comment is helpful. (Personally, I think your comment is a better answer than your answer - it boils down to "you could do that, but here are the challenges you'd have to overcome, and here's why it would be way more practical to just do it on Earth instead.") $\endgroup$
    – N. Virgo
    Jul 4 at 2:23
  • 1
    $\begingroup$ I'll look at working it in. $\endgroup$ Jul 4 at 2:28
10
$\begingroup$

Heat transfer in earth atmosphere

In earth atmosphere, heat is transferred from hot CPU to heat sink by conduction and then a fan transfers heat from heat sink to outside by convection (using air as medium for convection).

Heat transfer in space

As told here,

In a vacuum, heat can’t be transferred by conduction or convection. The process of convection and conduction requires some medium made of material particle for transmission of heat. In a vacuum, there is no material. So, heat travels in a vacuum by radiation.

In some situations, radiation may not be enough to transfer all the generated heat. Vacuum is a bad conductor of heat. In some setting, vacuum may increase heat.

$\endgroup$
4
  • 2
    $\begingroup$ This isn't the relevant comparison - you won't get down to 2.7K by convective cooling using a fan. $\endgroup$
    – N. Virgo
    Jul 3 at 22:06
  • 1
    $\begingroup$ Takes 12 to 26 hours for a human body to freeze in space for reference... Compare the two less than 1 hour for gas at the same temperature $\endgroup$ Jul 4 at 1:13
  • 2
    $\begingroup$ NASA's Lunar Reconnaisance Orbiter measured a temperature of -248 degrees Celsius (-415 degrees Fahrenheit) inside Hermite crater near the Moon's north pole, setting the record for the coldest-known place in our solar system and beating out even Pluto $\endgroup$ Jul 4 at 6:05
  • 2
    $\begingroup$ Quantum computers need to be cooled to extremely cold temperatures- Google's is cooled to just above absolute zero, about 0.01 kelvin. The bigger benefit of being earth-based is probably access to power, as the cooling apparatus requires 25 kilowatts of power. $\endgroup$
    – David
    Jul 5 at 22:50
8
$\begingroup$

Real spacecraft has a big problem with overheating, and must be equipped with cooling systems and radiators to prevent getting catastrophically hot. This is for devices that can operate at ~300K. A supercooled quantum computer would be very challenging.

It's very difficult to keep things cool in space. There's no air to blow over things to cool them down, which is how it's done on Earth. The laws of thermodynamics require that when you cool down an object, something else must be heating up more than you cool it down. Ask yourself, in the near vacuum of space, what will you be heating up and how? Other problems are that the Sun and reflected light from planets and moons will constantly shine on you and heat you up - even if you put up a shield, it will shine on the shield and heat it up, which will eventually heat up your entire spacecraft.

Some more logical locations for a QC are:

  • On a planet with a very distant orbit (like Pluto) and a thick atmosphere (just Earth-like is fine, but not super thin like Mars at 1% of Earth) near the poles.
  • On a liquid-covered planet or moon like Titan, where the liquid ocean can be used for powerful liquid cooling. The computer itself can be on a shore, an off-shore rig, suspended just under the surface from buoys or anchored to the ocean floor, these are all viable for heat management.
  • On a tidally locked planet, you can put the QC on the dark side (cold) or the twilight area (lots of natural wind).

Note that the idea is not to find an ambient that is cold enough for a QC, since that is unlikely. You're almost certainly going to have cooling systems, pumping heat from the very cold QC to much hotter outside. However, if the outside is not too much hotter (not hundreds of degrees), the coolant might be a bit more efficient.

$\endgroup$
3
  • $\begingroup$ Note that Pluto is actually covered in nitrogen ice. You're not going to get a thick atmosphere that far out that's anything but hydrogen/helium, and that pretty much takes a gas giant. Titan is actually close to the point of having its atmosphere start to condense, air cooling would work extremely well there even without the methane lakes. $\endgroup$ Jul 4 at 18:23
  • $\begingroup$ @ChristopherJamesHuff I didn't say pluto has a thick atmosphere. $\endgroup$ Jul 4 at 20:51
  • $\begingroup$ And my comment wasn't limited to Pluto. The reasons Pluto lacks a thick atmosphere will apply to anything less than a gas giant. Pluto, Triton, and everything in the Kuiper belt, like Sedna, Eris, Makemake, etc, are all too cold for the heavier atmospheric gases to be present in gas form as more than traces. Titan is about as far out as such a thing is possible, at a third the distance of Pluto. $\endgroup$ Jul 4 at 23:53
7
$\begingroup$

Space is not Cold

enter image description here

Put a cold thing in a vacuum flask and it stays cold a long time. Put a cold thing in the perfect vacuum of outer space and it stays cold forever.

Put a hot thing in a vacuum flask and it stays hot a long time. Put a hot thing in the perfect vacuum of outer space and it stays hot forever.

Heat is just molecules vibrating against one another. Hot things have the molecules vibrating faster.

Heat spreads because the molecules knock against each other. The hot thing cools because the molecules inside it knock against the air around it and lose vibration. The cold thing warms up because the molecules inside it knock against the air around it and gain vibration.

The vacuums work by having fewer molecules nearby the hot/cold thing. Fewer air molecules means less bumping and less spread of heat. No air molecules means no spread of heat.

The take-away is that a computer put in space stays at whatever temperature it was at launch.$^1$

Until you turn it on.

Then it starts creating heat. The heat stays inside the computer because of the vacuum. Space does not cool the computer. Space insulates the computer. It overheats. The end.


$ ^1$ It might absorb heat from the Sun. But that makes things harder and not easier.

$\endgroup$
12
  • 11
    $\begingroup$ "A hot thing [...] in the perfect vacuum of outer space and it stays hot forever": Not unless it has some internal heat source. It will radiate heat as black body radiation and if doesn't have an internal heat source it will eventually become very very cold. $\endgroup$
    – AlexP
    Jul 3 at 16:18
  • 5
    $\begingroup$ @JuimyTheHyena: If you have such good radiators then why would you bother going to outer space? They would work just as well here on Earth. $\endgroup$
    – AlexP
    Jul 3 at 16:19
  • 4
    $\begingroup$ "Put a hot thing in the perfect vacuum of outer space and it stays hot forever." Not really. Objects in space will lose thermal energy through radiative cooling. Your hot object is going to cool down over time... how long it takes depends on a bunch of factors. $\endgroup$
    – Corey
    Jul 4 at 2:06
  • 4
    $\begingroup$ nonono, this answer is wrong, space really is actually quite cold. First of all no real space is perfect vacuum second of all things will equilibrate (slowly, but not infinitely slowly) through electromagnetic radiation. When people refer to the temperature of space (usually referring to space far from the sun) they are usually referring to the temperature of the electomagnetic radiation which you are equilibrating with. $\endgroup$ Jul 4 at 7:35
  • 2
    $\begingroup$ "Put a hot thing in ... outer space and it stays hot forever": The earth is in space, and it stays at roughly the same temperature despite being constantly bombarded by energy from the sun. Why doesn't its temperature constantly increase? Because it is also constantly radiating away its heat, mostly in the infrared. Would the earth stay at its current temperature forever if it were somehow ejected from the solar system? Of course not. It would freeze. Any object in space, shielded from (or distant from) sources of radiation, will do the same. Search for voyager temperature. $\endgroup$
    – phoog
    Jul 4 at 12:19
6
$\begingroup$

The "natural coldness" is an oversimplification. The lack of convection complicates temperature management. Couple that with the maintenance issues, and it sounds like a very bad idea.

$\endgroup$
1
  • 1
    $\begingroup$ Ow you need "arrived" technology to put something in space. Not a complicated experimental machine you'd like to be able to maintain and service. But the opening has not stated technology level, it could be far future when "large scale" capabilities are needed. Maybe future quantum computers are standard, closed box building blocks that come with a standard cooling module. Excess heat could be used for propulsion.. $\endgroup$
    – Goodies
    Jul 3 at 12:31
4
$\begingroup$

Space isn't cold... it's empty.

I'm building on a number of answers here and would like to point out that the answer you selected as the "best answer"... isn't... (all due respect to @ChristopherJamesHuff, who kinda answered the question but didn't adequately explain anything).

Your problem isn't heat. Your problem is the transfer of heat away from what's generating it. If you don't do that fast enough, the object generating heat (your computer) burns up.

On Earth we have gases (usually the atmosphere) and liquids (you should be thinking "radiator fluid") to do that for us. But even that's often complicated. Let's consider your car!

  • Your engine (OK, the engine in my car...) is burning a not-quite-as-aptly-named-as-we-once-thought fossil fuel. Nothing's perfect, and as a result the tiny explosion that moves a piston also generates heat. We don't want the heat. In an ideal world there wouldn't be any heat. A massive amount of electronic engineering is dedicated to minimizing the generation of heat! But there's heat. What's next?

  • Your engine block has tunnels in it that allow the passage of a fluid — radiator fluid, to be specific. This fluid and the tunnels in the engine block are specifically designed to transfer as much heat to the fluid as is humanly possible as quickly as humanly possible. But this isn't enough! There's a finite amount of radiator fluid and although it's great at sucking the heat out of the engine block, it can only absorb so much before it first boils, then vaporizes, causing all kinds of trouble.

  • So the next thing that happens is that fluid runs through a radiator that allows the heat to move from the fluid to the atmosphere. It's technically true that the atmosphere is finite... but compared to radiator fluid, it's infinite and can absorb an infinite amount of heat.1

To make my point, we moved the heat from the source (combustion) into the engine block, then into radiator fluid, then into the atmosphere.

The problem with space is that it's empty

No atmosphere. Certainly no fluids. There's nowhere for the heat to go.

Well... Kinda...

A computer on the ground that needs a couple of cubic inches of heat sink to vent the heat to Earth's atmosphere needs cubic feet of heat sink to vent the heat to the precious few atoms that can absorb it in space. That's the problem. There's pretty much nowhere for the heat to go.

Why, then, do we think space is cold? Because where there is nothing to hold heat, the result is the perception of cold. Why the perception of cold?

This is really important!

The nature of the universe (all of the universe) is to move to the lowest energy state. We call this "entropy." Things don't want to be hot. Energy wants to even out and stabilize. When you take away the source of the heat, the result (over time) is that everything falls to a minimum temperature. We call that "cold" because we're conditioned to living under the beauty of a blazing sun.

But the truth is, space isn't so much cold as it is empty. In the middle of all that empty the energy, through entropy, has minimized and the result is "cold." But that doesn't mean you can sink heat into it. After all, if the blazing glory of the sun can't heat up space, why should your itsy-bitsy computer?

Conclusion

It would be better to put your computer on Antarctica where there's both a lot of cold and a lot of gasses and fluids to draw away the heat. But if you insist on space, you'll need massive heat sinks to transfer the heat to what little mass there is out there.

Heat must be transferred through something. In a perfect vacuum (which space isn't, thankfully!) there isn't anything heat can transfer through.


1This isn't actually true. While a minor component of climate change compared to polluting the atmosphere with greenhouse gasses, the heat generated through combustion is nevertheless part of the problem. The atmosphere can only hold so much heat, too, without consequences. Like melting ice caps.

$\endgroup$
1
  • $\begingroup$ Your answer completely ignores radiative heat transfer. Space has nothing that can be used for heat conduction or convection. but there is still blackbody radiation, which can both add heat (from nearby sources) and take it away. Spacecraft require massive "heat sinks" (properly known as radiators) because radiation is slow, not because it's using "what little mass there is out there." It's not using mass to transfer heat, it's using photons. $\endgroup$
    – Brianorca
    Jul 5 at 20:50
1
$\begingroup$

You'd have to use geothermal cooling, i.e. shadowed craters on the moon have a constant temperature of -240 to -245.

If you have nuclear energy there to go near -273 it would be good.

It would be a big rig. Very complicated. Then you could decrypt SHA 256bit codes.

$\endgroup$
1
$\begingroup$

And most importantly, would the smaller temperature gap make a real difference?

No, as pointed out in the other answers about the coldness of space. Even if space was somehow magically able to cool things down, anything you do in space can be done on Earth cheaper, faster, and larger.

Do not under estimate how expensive it is to ship material to space. On Earth? You get a quote from a shipping company and move on with your day without worrying whether your moving 1 tonne or 1.001 tonnes of material. In space that is potentially needing you to move to a new rocket or accept a shorter mission as you put less consumables in the device to save mass.

Turns out you want a bigger machine than originally planned? On Earth you combine two rooms into one. In space? Your looking at at least launching an entirely new satellite.

What happens when the engineers turn round and say they need a few hundred extra watts of power to run things? Or an extra few kilograms of support structure? On Earth, who cares. In space, well you've looking at a large redesign to save mass/power elsewhere.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .