• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Data center in Space

There's no way it's cost effective. Remote hands is gonna cost a fortune.

Seriously though, power is hard, cooling is hard, networking is hard, maintenance is impossible, and the thing will burn up in a few years. More bitflips up there, too. At least you can get a couple bucks from liquidating 3-5 year old servers on earth.

I guess data destruction is easier though.
 
Sounds like a total pipe dream. In addition to the very high cost of getting things in space (about $1000/pound minimum these days), there the biggies of power and cooling. Even in a high orbit where you can stay always in the sun, solar just takes a lot of panel to produce much power. That's why when you see most satellites or things like the Hubble there's quite a lot of panels compared to the total size, even though they are usually not that high power. They would need a shit ton to make a modern data center work, even a small one. I guess they could try a nuclear battery instead, not sure what the output on those are. Cooling is another problem, while it is very cold in low earth orbit, there's also very little gas so it takes a long time to conduct heat away. Again, big issue for modern DCs and any kind of active cooling is even MORE weight and MORE power that is needed.

I really have trouble seeing how the costs would be at all worth it. Like what does it get you to have a datacenter in space, rather than on land? About the only answer I ever hear is the tech-bro "You wouldn't be subject to any laws man!" But of course that is false, the companies and the people who run them still very much live on Earth and thus don't escape any more than putting a datacenter in another country would.

The problem we have with the massive number of datacenters being built isn't that we lack the land, they don't really use that much compared to a lot of things, the issue is power and cooling (water use for the cooling in particular). There are more minor issues of noise and such but those would be easily solved building them out in the middle of nowhere, which would still be way more accessible than in space.
 
Wont ever happen, total brain dead idea..

The latency alone to move data back and forth kills it, also since there is literally no air to cool them, the size of surface you would need to dissipate ANY heat..

I will need to find a write up a person did of all the basic laws of physics and other crap that would not make this work at all.
 
Sycraft said:
They would need a shit ton to make a modern data center work, even a small one. I guess they could try a nuclear battery instead, not sure what the output on those are. Cooling is another problem, while it is very cold in low earth orbit, there's also very little gas so it takes a long time to conduct heat away. Again, big issue for modern DCs and any kind of active cooling is even MORE weight and MORE power that is needed.
Click to expand...

Well you don't conduct heat away in space you use radiators, which can be greatly effective (Yey T^4 values for radiative heat transfer), although you also have to worry about taking in heat from basically everything in existence with a LoS too, but like any cooling stuff requires a good amount of surface area, which you'd probably use some sort of active cooling to get the heat from the electronics throughout the radiators. I guess there's no reason you couldn't use loop or oscillating heat for that. There's just a lot of annoyances with cooling things in space, I did one of my grad school Design of Thermo systems projects on designing a cooling system for a small satellite, and it definitely isn't fun. Calculating the view factors alone becomes a fun nightmare of PDEs . Not terrible with some numerical analysis solvers though, but bleh.

While I don't see DCs in space, I do see, and it makes for some interesting use cases for Edge computing in space. Even throwing additional compute out there to have it.
 
Balkroth said:
Well you don't conduct heat away in space you use radiators, which can be greatly effective (Yey T^4 values for radiative heat transfer), although you also have to worry about taking in heat from basically everything in existence with a LoS too, but like any cooling stuff requires a good amount of surface area, which you'd probably use some sort of active cooling to get the heat from the electronics throughout the radiators. I guess there's no reason you couldn't use loop or oscillating heat for that. There's just a lot of annoyances with cooling things in space, I did one of my grad school Design of Thermo systems projects on designing a cooling system for a small satellite, and it definitely isn't fun. Calculating the view factors alone becomes a fun nightmare of PDEs . Not terrible with some numerical analysis solvers though, but bleh.

While I don't see DCs in space, I do see, and it makes for some interesting use cases for Edge computing in space.
Click to expand...
I'm not saying you can't cool things in space, clearly we can, the JWST is a great example requiring a very low operating temperature for the sensor to work right. The problem is enough cooling for the servers. Consider that something like a DGX H200 (nVidia GPU system often used for AI shit) can draw as much as 10kW in a 8U rack space. That's a lot of power in a small space. How do you deal with cooling that, or let's say a full rack, which is 5 of them?

I just don't see the use case. Sure devices themselves will have computing power in them, they already do and that only increases, but to just put a container full of computers in orbit? Why? So many problems to be solved, all of which are much simpler on land and we don't lack for land.
 
Sycraft said:
I'm not saying you can't cool things in space, clearly we can, the JWST is a great example requiring a very low operating temperature for the sensor to work right. The problem is enough cooling for the servers. Consider that something like a DGX H200 (nVidia GPU system often used for AI shit) can draw as much as 10kW in a 8U rack space. That's a lot of power in a small space. How do you deal with cooling that, or let's say a full rack, which is 5 of them?

I just don't see the use case. Sure devices themselves will have computing power in them, they already do and that only increases, but to just put a container full of computers in orbit? Why? So many problems to be solved, all of which are much simpler on land and we don't lack for land.
Click to expand...
Oh I'm in complete agreement, that is a ton of energy to both generate and dissipate, heck we have trouble cooling them on the ground, (yey for the 80+kw racks we're seeing). I'd imagine that's where Asics come in for lighter load. Still doesn't really make a ton of sense, at least currently. It's probably a decent thing to POC out though just to see how things work.


Full data centers though, yea, not really. Honestly we should be throwing up DCs in places like the Nordic countries where they have cold weather geothermal energy pretty readily available to tap.
 
Starlink already has much of the infrastructure in place. They already have hardware delivery, they have power, they have networking, and they have a replacement schedule. Although a 1,500 watt GPU is pretty intense for a satellite, Starlink will have the head start with quantity. 10,000 inter-networked computers up in space talking to each other has potential written all over it. What the potential is, no one knows yet, but money prone people are coming up with ideas.
 
Two really big problems with this. First is radiation hardening, without the atmosphere those tiny semiconductors with get turned into useless Swiss cheese in short order. Second, cooling, any data center will have a limited thermal mass, and the only way to cool past that is radiating energy away which isn't exactly that efficient
 
sfsuphysics said:
Second, cooling, any data center will have a limited thermal mass, and the only way to cool past that is radiating energy away which isn't exactly that efficient
Click to expand...
They could require good (better than now..): https://en.wikipedia.org/wiki/Liquid_droplet_radiator ?

North of Canada (or like in Finland place that can use the datacenter heat in their already installed city wide heating system), do seem to take care of a lot of issues with datacenter in a simple way.

Sycraft said:
Why?
Click to expand...
That would be how attractive solar energy in space could get and how important of a variable it would be, lower latency for some things being the second one.
 
R.jpg
 
toast0 said:
There's no way it's cost effective. Remote hands is gonna cost a fortune.

Seriously though, power is hard, cooling is hard, networking is hard, maintenance is impossible, and the thing will burn up in a few years. More bitflips up there, too. At least you can get a couple bucks from liquidating 3-5 year old servers on earth.

I guess data destruction is easier though.
Click to expand...
I don't know.... Microsoft's Project Natick was a resounding success, and that stuck a datacenter in a sealed container under the ocean.
Putting them in low Earth orbit seems like the next logical step if you look at things from a "wouldn't this be freeking awesome!" perspective, remembering that they have more money than they know what to do with.
 
Lakados said:
I don't know.... Microsoft's Project Natick was a resounding success, and that stuck a datacenter in a sealed container under the ocean.
Putting them in low Earth orbit seems like the next logical step if you look at things from a "wouldn't this be freeking awesome!" perspective, remembering that they have more money than they know what to do with.
Click to expand...
very very very different, easy to get power and low latency networking to it, as well as maintain it.

What is the cost if a system fails, goes down, what ever and time to recovery, imagine the redundancy that would be needed. Sure Starlink is great but when you have 10's of GB's of data needing moving around with little latency.....

It is outright a stupid idea.
 
maybe once they get the AI data center into orbit they can send it out into space. until it comes back to destroy all humans.
 
MrGuvernment said:
very very very different, easy to get power and low latency networking to it, as well as maintain it.

What is the cost if a system fails, goes down, what ever and time to recovery, imagine the redundancy that would be needed. Sure Starlink is great but when you have 10's of GB's of data needing moving around with little latency.....

It is outright a stupid idea.
Click to expand...
The core maintenance philosophy of Project Natick was to eliminate on-site physical maintenance entirely due to the difficulty of accessing a submerged data center.
  • "Lights-out" operation: The data centers were sealed in a dry nitrogen atmosphere to prevent corrosion and were designed to run autonomously.
  • Fail-in-place: Instead of repairing or replacing components as they failed (which is standard practice in land-based data centers), the system was designed with enough redundancy that failed servers or disks were simply taken offline, and operations continued without disruption.
  • Periodic overhaul: The entire data center unit was designed to be retrieved from the seafloor every five years, at which point all servers would be swapped out and recycled, and the module redeployed.
What Microsoft discovered at the end of the project was that they experienced 1/8'th the failure rate of hardware in that environment that they expected based on similar hardware failure rates in their traditional datacenters.

Solar power works exceptionally well in low Earth orbit, and advancements in satellite tech have made high-speed speed low-latency networking a cheap thing to do in that environment. StarLink works by taking multiple small dishes and simulating one large one through some software black magic. It can be scaled up incredibly well, especially if you aren't worried about the physical footprint of the ground unit.
Microsoft already operates Datacenters that cost it upwards of $1.3B a year to operate, and that is utilizing sweetheart electrical deals and exemptions on land taxes; if those deals end, that number could easily double, at which point they are looking at numbers rivaling the International Space Station in terms of annual upkeep.
So if you are looking to host a large AI infrastructure, but you don't have the same level of Government in's that the likes of Microsoft and Alphabet do, then you are in for a very expensive world of hurt.

Project Natick proved that it could work in concept, and Solar advancements have made for lighter, more efficient panels, which could greatly assist in power delivery.


You wouldn't need to build something the size of a large platform in space with living quarters and the like. It could be a series of self-contained units which wouldn't be much larger than your traditional communications satellite, which already costs around $400 million a pop.
 
Lakados said:
The core maintenance philosophy of Project Natick was to eliminate on-site physical maintenance entirely due to the difficulty of accessing a submerged data center.
  • "Lights-out" operation: The data centers were sealed in a dry nitrogen atmosphere to prevent corrosion and were designed to run autonomously.
  • Fail-in-place: Instead of repairing or replacing components as they failed (which is standard practice in land-based data centers), the system was designed with enough redundancy that failed servers or disks were simply taken offline, and operations continued without disruption.
  • Periodic overhaul: The entire data center unit was designed to be retrieved from the seafloor every five years, at which point all servers would be swapped out and recycled, and the module redeployed.
What Microsoft discovered at the end of the project was that they experienced 1/8'th the failure rate of hardware in that environment that they expected based on similar hardware failure rates in their traditional datacenters.

Solar power works exceptionally well in low Earth orbit, and advancements in satellite tech have made high-speed speed low-latency networking a cheap thing to do in that environment. StarLink works by taking multiple small dishes and simulating one large one through some software black magic. It can be scaled up incredibly well, especially if you aren't worried about the physical footprint of the ground unit.
Microsoft already operates Datacenters that cost it upwards of $1.3B a year to operate, and that is utilizing sweetheart electrical deals and exemptions on land taxes; if those deals end, that number could easily double, at which point they are looking at numbers rivaling the International Space Station in terms of annual upkeep.
So if you are looking to host a large AI infrastructure, but you don't have the same level of Government in's that the likes of Microsoft and Alphabet do, then you are in for a very expensive world of hurt.

Project Natick proved that it could work in concept, and Solar advancements have made for lighter, more efficient panels, which could greatly assist in power delivery.


You wouldn't need to build something the size of a large platform in space with living quarters and the like. It could be a series of self-contained units which wouldn't be much larger than your traditional communications satellite, which already costs around $400 million a pop.
Click to expand...
Cooling...and ionized particles this is a pipedream water cools better than air...a vacuum does not.
Combine that with launch costs and you have the recipe for total failur (and a non-working business plan) but if you think this is "clever" i have a brigde for sale and a plot of land on Mars.
 
The coldness of space will handle any heat dissipation but I'm not sold on 'wireless/WiFi' data centers.

Latency aside, a lot can go wrong even in low Earth orbit with radiation, specifically for datacenter type 'critical' usage. Not like just trying to send a radio transmissions to talk to some astronauts or watching YouTube via StarLink.
 
  • Like
Reactions: erek
like this
data center in space likely to fail harder than data centers in the ocean

1766089212790.jpeg
 
littlegreenmenfromhome said:
Cooling...and ionized particles this is a pipedream water cools better than air...a vacuum does not.
Click to expand...

The real problem with a space datacenter comes whenever in sunlight and then being bombarded with all that additional energy/radiation/heat. If you attach radiators/heatsink fins to a big container/datacenter and left it in complete darkness to drift/orbit, heat would leave as the object moves (replacing air movement - the movement doesn't cause the cooling just keeps it away from any heat already dissipated though really might not be an issue there because a vacuum) even though it's in a vacuum.
 
Last edited:
erek said:
data center in space likely to fail harder than data centers in the ocean

View attachment 773453
Click to expand...
It didn't fail....
Project Natick was a technical success, proving underwater data centers are reliable, sustainable, and efficient, with lower failure rates than land-based ones due to constant cool temperatures and nitrogen. However, Microsoft ended the project in 2024, deciding it wasn't commercially viable, pivoting to apply its lessons in robotics and sustainable infrastructure to other areas, even as other companies explore similar concepts.
Successes of Project Natick:
High Reliability: Servers failed less often underwater (1 in 8) than on land (8 in 135) due to stable temperatures and inert gas (nitrogen), avoiding oxygen's corrosive effects.
Energy Efficiency: The cold ocean water provided natural cooling, reducing the energy needed for climate control.
Sustainability: The test pods used renewable energy (Orkney's wind/solar) and were designed to function without human intervention, reducing land use.
Feasibility Proven: It demonstrated that deploying and operating data centers underwater is possible, even in harsh conditions.
Why it ended:
Operational Challenges: While technically sound, the cost and logistical hurdles of building, deploying, and retrieving these massive underwater units outweighed the benefits for Microsoft.
Shift in Focus: Microsoft is now exploring new tech like advanced robotics for land-based data centers and novel power sources, applying Natick's insights to broader sustainability goals.

Microsoft gets a lot of sweetheart deals for power and exemptions from many land taxes that don't require them to take these sorts of steps to circumvent them, but most other companies don't, so for them it could be an atractive alternative.
 
Lakados said:
It didn't fail....
Project Natick was a technical success, proving underwater data centers are reliable, sustainable, and efficient, with lower failure rates than land-based ones due to constant cool temperatures and nitrogen. However, Microsoft ended the project in 2024, deciding it wasn't commercially viable, pivoting to apply its lessons in robotics and sustainable infrastructure to other areas, even as other companies explore similar concepts.
Successes of Project Natick:
High Reliability: Servers failed less often underwater (1 in 8) than on land (8 in 135) due to stable temperatures and inert gas (nitrogen), avoiding oxygen's corrosive effects.
Energy Efficiency: The cold ocean water provided natural cooling, reducing the energy needed for climate control.
Sustainability: The test pods used renewable energy (Orkney's wind/solar) and were designed to function without human intervention, reducing land use.
Feasibility Proven: It demonstrated that deploying and operating data centers underwater is possible, even in harsh conditions.
Why it ended:
Operational Challenges: While technically sound, the cost and logistical hurdles of building, deploying, and retrieving these massive underwater units outweighed the benefits for Microsoft.
Shift in Focus: Microsoft is now exploring new tech like advanced robotics for land-based data centers and novel power sources, applying Natick's insights to broader sustainability goals.

Microsoft gets a lot of sweetheart deals for power and exemptions from many land taxes that don't require them to take these sorts of steps to circumvent them, but most other companies don't, so for them it could be an atractive alternative.
Click to expand...
i know it’s a “technical” success

but ocean-based data centers definitely didn’t kick off a mass market trend or adoption

it’s a genuine failure even by your own quote, “Operational Challenges: While technically sound, the cost and logistical hurdles of building, deploying, and retrieving these massive underwater units outweighed the benefits for Microsoft.
Shift in Focus: Microsoft is now exploring new tech like advanced robotics for land-based data centers and novel power sources, applying Natick's insights to broader sustainability goals.”
 
Sycraft said:
I really have trouble seeing how the costs would be at all worth it. Like what does it get you to have a datacenter in space, rather than on land? About the only answer I ever hear is the tech-bro "You wouldn't be subject to any laws man!" But of course that is false, the companies and the people who run them still very much live on Earth and thus don't escape any more than putting a datacenter in another country would.
Click to expand...

Even if you're subject to laws, how would they know what exactly your men are doing up there? I guess the only way is to have strict surveillance on what goes up there, but once it gets up there, it's not like we have international space police. Well, the saving grace is that this is probably not viable with our current tech anyway. The only way I can see it being somewhat viable because they can technically capture 100% of the heat emissions in a pressurized space environment, so I guess supposedly you could vent that into a pressurized heat chamber to reuse the power that's being wasted via heat back into more electricity...? But the only source of power you have is solar panels. There's a decent amount of debris up there. Build anything to decent size and surface area, and the probability of it being struck raises exponentially, which would need repairs. Repairs are expensive, unless they have AI repair bots up there (or maybe remote controlled space drones, which would probably need refueling at some point).

Also the only way to get data back from it would be via satellite. Which has come quite a long ways, but long enough to satisfy demand?

Lakados said:
Energy Efficiency: The cold ocean water provided natural cooling, reducing the energy needed for climate control.
Click to expand...
Meanwhile, all the nearby dead fish due to a tiny temperature increase:
 
erek said:
i know it’s a “technical” success

but ocean-based data centers definitely didn’t kick off a mass market trend or adoption

it’s a genuine failure even by your own quote, “Operational Challenges: While technically sound, the cost and logistical hurdles of building, deploying, and retrieving these massive underwater units outweighed the benefits for Microsoft.
Shift in Focus: Microsoft is now exploring new tech like advanced robotics for land-based data centers and novel power sources, applying Natick's insights to broader sustainability goals.”
Click to expand...
Yeah, but the operational wording there is "For Microsoft"

Microsoft pays exceptionally low prices for the electricity it uses for its data centers, which is covered over here.
https://hardforum.com/threads/senat...rs-pass-energy-costs-on-to-americans.2045485/

Combine that with the creative accounting (which they are now in numerous legal battles about), which gives them exceptionally low property taxes and land usage taxes around the world, and Microsoft doesn't need to worry about the physical size of their datacenters in the same ways that anybody else needs to.

But anybody new to the space absolutely does need to worry about that, so for them, underwater and low-earth orbit are attractive alternatives.
 
StoleMyOwnCar said:
Even if you're subject to laws, how would they know what exactly your men are doing up there? I guess the only way is to have strict surveillance on what goes up there, but once it gets up there, it's not like we have international space police.
Click to expand...
I mean, to be useful it would need to communicate with things on Earth. No point in having a datacenter that doesn't communicate. So if you are using it to host your illegal files, well then you are still on the hook for hosting them. Again, you don't really gain anything over trying to host it in a different country.
 
Lakados said:
I don't know.... Microsoft's Project Natick was a resounding success, and that stuck a datacenter in a sealed container under the ocean.
Putting them in low Earth orbit seems like the next logical step if you look at things from a "wouldn't this be freeking awesome!" perspective, remembering that they have more money than they know what to do with.
Click to expand...
I'm not saying you can't make a (small) datacenter in space; or underwater. I'm just saying it's not a good idea, because it makes everything hard and costly. Microsoft doesn't run any underwater servers, because it doesn't make sense to do it beyond research, even though you can.

Space datacenter for research --- sure; constraints drive creativity, and that's good for things. Space datacenter for serving actually terrestrial needs? No, there's no good reason. Maybe you could stuff some storage and host CDN nodes in starlink, although you're probably going to figure out how to get the top CDNs to collaborate on a shared system cause you don't really want one box from each major provider.
 
Lakados said:
Yeah, but the operational wording there is "For Microsoft"

Microsoft pays exceptionally low prices for the electricity it uses for its data centers, which is covered over here.
https://hardforum.com/threads/senat...rs-pass-energy-costs-on-to-americans.2045485/

Combine that with the creative accounting (which they are now in numerous legal battles about), which gives them exceptionally low property taxes and land usage taxes around the world, and Microsoft doesn't need to worry about the physical size of their datacenters in the same ways that anybody else needs to.

But anybody new to the space absolutely does need to worry about that, so for them, underwater and low-earth orbit are attractive alternatives.
Click to expand...
it’s a failure regardless of technically feasible or. not


the economics don’t justify it by your own quote and even microsoft has pivoted
 
It would be interesting to see if the Tensor chips can contend with the riggers of operating in those conditions. And maybe someone up at those space stations can finally play Genshin Impact without boiling their hands off.
 
Lakados said:
Actually, the opposite, they found it created an artificial reef and was a net positive. They filed a patent on it because of the results.
Click to expand...

I think that depends on what a "net positive" means. I have a fish tank and let me tell you those little shits die without much changing. Now, out in the ocean I assume most temperature change is going to be very contained, especially if the scale of the operation was very small, but I would really be surprised if some fish didn't die because of it... somehow.

Frankly though... I think the idea might need to come back, considering how much these fucking datacenters are messing everything up. Making them a bit more efficient while forcing companies to employ people to put them down there and do maintenance on them? Yeah sounds like a net positive to me. For once AI would actually be creating some jobs, rather than destroying everything while everyone sits here shooting the breeze about all of the hypothetical jobs that it's supposedly going to create that don't look like they're going to exist any time soon. It would also stop taking up space on land. Maybe Microsoft needs to reintroduce it as a "way to get some positive PR with the environmentalist people" and "tax writeoffs".

Sycraft said:
I mean, to be useful it would need to communicate with things on Earth. No point in having a datacenter that doesn't communicate. So if you are using it to host your illegal files, well then you are still on the hook for hosting them. Again, you don't really gain anything over trying to host it in a different country.
Click to expand...
The question is how is anyone gonna find out, I guess? And frankly they could do anything up there on a space station. "Storing illegal files" would be the easiest one. But let's say we go with this scenario and someone claimed they're hosting illegal files... how would they go about finding out and charging them? Or making sure they're gone? It would be pretty expensive for a government to enforce compared to just storming your data center on the ground and going "show us this shit". Frankly one application is they could just pirate training data, beam it up, and then all that comes out is a model. Obfuscate the returned lines a bit, and who's going to storm your space station to see if it used pirated training data? Just secure delete it on the ground after you're done, and it's a done deal.

Well not that it matters, we already have companies pirating training data and literally getting away with it in court...
 
toast0 said:
I'm not saying you can't make a (small) datacenter in space; or underwater. I'm just saying it's not a good idea, because it makes everything hard and costly. Microsoft doesn't run any underwater servers, because it doesn't make sense to do it beyond research, even though you can.

Space datacenter for research --- sure; constraints drive creativity, and that's good for things. Space datacenter for serving actually terrestrial needs? No, there's no good reason. Maybe you could stuff some storage and host CDN nodes in starlink, although you're probably going to figure out how to get the top CDNs to collaborate on a shared system cause you don't really want one box from each major provider.
Click to expand...
Depending on what electricity and land taxes cost, but yeah.
 
If you could solve the wireless transmission problem so that latency, speed and reliability aren't an issue under any workloads/necessity of uptime - it would probably/might be better to build datacenters on the close side of the moon. Being on the close side would help with transmission direction/beaming - the far side can act as a shield/barrier to debris/asteroids/etc for (some sides of) your datacenter vs worrying about all sides all the time if 'freestanding/in orbit' - if you're already building stuff on the moon you can add or already have bases for personnel which would negate some of the 'remote repair problem'.

You'd need to build an awning of sorts over the structure that could reflect sunlight so it doesn't hit/get absorbed by the exterior/heatsink (and could also act as additional shield/barrier to other sides of the structure from debris); and also heat could be used to melt ice on/in the moon for either personnel (drinking) or for cooling, maybe help with some power generation too (but you need -maybe more than the moon provides - gravity to make the water naturally 'flow' for that to generate your power, otherwise you waste power trying to make it do that, so maybe not power - if generating steam power with the water maybe)?

Also the moon has a slight but still existing atmosphere, I don't know how much that would/could affect dissipating heat out into the vastness and coldness of space/the vacuum 🤔
 
Last edited:
kram182 said:
If you could solve the wireless transmission problem so that latency, speed and reliability aren't an issue under any workloads/necessity of uptime - it would probably/might be better to build datacenters on the close side of the moon. Being on the close side would help with transmission direction/beaming - the far side can act as a shield/barrier to debris/asteroids/etc for (some sides of) your datacenter vs worrying about all sides all the time if 'freestanding/in orbit' - if you're already building stuff on the moon you can add or already have bases for personnel which would negate some of the 'remote repair problem'.

You'd need to build an awning of sorts over the structure that could reflect sunlight so it doesn't hit/get absorbed by the exterior/heatsink (and could also act as additional shield/barrier to other sides of the structure from debris); and also heat could be used to melt ice on/in the moon for either personnel (drinking) or for cooling, maybe help with some power generation too (but you need -maybe more than the moon provides - gravity to make the water naturally 'flow' for that to generate your power, otherwise you waste power trying to make it do that, so maybe not power)?

Also the moon has a slight but still existing atmosphere, I don't know how much that would/could affect dissipating heat out into the vastness and coldness of space/the vacuum 🤔
Click to expand...
The side of the moon we can see reaches temperatures of up to 121°C (250°F), but the lack of moisture in the limited atmosphere means temperature changes very quickly in a short distance, so under a sun shade within a foot of the temperatures could dip as low as -133°C (-208°F).
But it would be much easier instead to make use of the underground cave system that exists on the moon from the ancient lavaflows from when it was formed.

But interestingly, between the 2024 Artemis Accords, the 1979 Moon Agreement, and the 1967 Outer Space Treaty, any private company attempting to build such an installation on the moon could need as many as 139 countries to approve of the plan beforehand.
 
Lakados said:
But interestingly, between the 2024 Artemis Accords, the 1979 Moon Agreement, and the 1967 Outer Space Treaty, any private company attempting to build such an installation on the moon could need as many as 139 countries to approve of the plan beforehand.
Click to expand...

It's a good thing the corpos run the show and could just pay anyone here on Earth off - but could they pay off the aliens on the far side or inside the moon? 👽🔫😁🤔
 
This might be needed reading:
https://en.wikipedia.org/wiki/Spacecraft_thermal_control

"Most spacecraft radiators reject between 100 and 350 W of internally generated electronics waste heat per square meter. Radiators' weight typically varies from almost nothing, if an existing structural panel is used as a radiator, to around 12 kg/m2 for a heavy deployable radiator and its support structure."

This thread is a prime example of people not understanding physics.
 
kram182 said:
The coldness of space will handle any heat dissipation but I'm not sold on 'wireless/WiFi' data centers.

Latency aside, a lot can go wrong even in low Earth orbit with radiation, specifically for datacenter type 'critical' usage. Not like just trying to send a radio transmissions to talk to some astronauts or watching YouTube via StarLink.
Click to expand...
No it will not.
A vacuum is a space with no air, meaning no molecules to carry heat via conduction or convection this is the primary ways heat moves on Earth without air heat escapes only through radiation (infrared light).

A vacuum is what help insulate multi layered windows.

Physics understanding is required, this is not some TV show "Sci-Fi".
 
You must log in or register to reply here.
Back
Top