Hacker News Reader

Comment by delichon | Hacker News Reader

NIMBY for data centers is opportunity for SpaceX. When they saturate the demand for communication, data processing demand will be ramping up with no apparent ceiling. The merger between SpaceX and xAI positions them to benefit both from the AI revolution, and from the resistance to it. It's like a hypothetical 19th century textile company that managed to profit from Luddite riots by using them to help move production to Umpa Loompa.

View on HN

Space doesn't seem like a good place to build datacenters at all. Cooling is going to be an enormous issue, how do you disperse of heat in a vaccuum? Radiators are very ineffective for cooling.

Cooling is the biggest reason for space datacenters, heat is movement of particles and vacuum being an absence of particles, so it's by definition cold.

Naturally the system needs energy, the sun giving radiation convertible to electricity should enable that.

Both parts are documented about ISS

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

Space is indeed very cold, but it does not cool you off very quickly. The lack of particles means it's harder to get heat away from yourself. Essentially all of the energy produced via solar panels would be converted to heat by the computers.

It's cold in a sense that is not very relevant. Your tumbler has a vaccuum layer because vaccuum does not transport or absorb any heat. you need those atoms to carry away heat.

> Cooling is the biggest reason for space datacenters, heat is movement of particles and vacuum being an absence of particles, so it's by definition cold.

That absence of particles also means it is a nearly perfect insulator, with almost no heat transferred from the station to space by contact with space. No convection either.

That leaves cooling by thermal radiation, which is not a very good method.

Do data centers in space actually make sense? I can’t figure out how that’s possible. But some people seem to believe they do(?)

> Do data centers in space actually make sense?

No. It's currently a fantasy. Even if the cost of getting payloads to orbit decreased another x100, you still have the issues of radiation and heat dissipation.

this will age poorly. you have both Google, Tesla/X betting on it. They are not stupid and probably have given it way more thought than people's whose paycheques not tied to this have thought about.

This is an ambitious bet, with some possibility of failure but it should say a lot that these companies are investing in them.

I wonder what people think, are these companies so naive?

Edit: Elon, Sundar, Jensen, Jeff are all interested in this. Even China is.

What conspiracy is going on here to explain it? Why would they all put money into this if it is so obvious to all of you that it is not going to work?

Google is hardly betting on it; they are exploring the feasibility of it and are frank about the engineering challenges: > significant engineering challenges remain, such as thermal management, high-bandwidth ground communications, and on-orbit system reliability.[1]

[1] https://research.google/blog/exploring-a-space-based-scalabl...

why do you think this changes what i said? I know it has constraints but the fact is that Google is serious about it. Enough to publicly speak about it many times and invest enormous amounts of R&D.

You are saying they are "hardly betting on it". This is grossly false and I wonder why you would write that? Its clearly a serious bet, with lots of people working on it.

> Google CEO Sundar Pichai says we’re just a decade away from a new normal of extraterrestrial data centers

Its surely a high risk bet but that's how Google has been operating for a while. But why would you say they are hardly betting on it?

As a counter question: do you think Google is not serious about it?

I never said Google wasn't serious; I said they are hardly betting on it relative to their other capital expenditures. Google rightfully describes this as a "moonshot." To date, the only public hardware commitment is two prototype satellites in 2027 for a feasibility study. Compared to the billions pouring into Waymo, DeepMind, and terrestrial data centers, this doesn't yet qualify as an "enormous" financial bet, even if the engineering intent is serious.

i agree with you then. lets agree that the intent is serious.

Everyone were also betting on quantum computing and the hydrogen energy revolution.

My napkin math says that, for a system at around 75°C, you would need about 13,000 square meters of radiators in space to reject 10 MW of heat.

why do you think they are betting on it if it so obvious to you that it won't work?

Just because some CEOs pour billions into fantasy projects it doesn't mean they're viable. Otherwise we all would be in the metaverse wouldn't we?

Sure, I don't claim all of them go well. Do you want to run a hypothetical exercise on how many they get right vs wrong? And based on that we can see if this is a "fantasy" or not?

No but you're claiming "if they all are investing X amount then these bets obviously must pan out". If you follow that rationale then it means that all bets that these company's make in the same space must all pan out. So if they don't all pan out then the fact that they're all making bets isn't a sound rationale for it being true.

As others have pointed out, investors notoriously have FOMO, so rationale actors (CEOs of big tech) naturally are incentivized to make bets and claims that they are betting on things that the market believes to be true regardless if they are so as to appease shareholders.

No, i don't claim that all bets must pan out - simply that most bets are made intelligently with serious intent.

that's the way i see this bet as well.

your take on investors is naive and largely incorrect - its the musical chairs theory of markets.

> simply that most bets are made intelligently with serious intent.

That is NOT what you said. You said this:

> Why would they all put money into this if it is so obvious to all of you that it is not going to work?

In other words: "if these companies are putting all of this money to work then it's obvious it will work"

So, no you didn't simply say "their bets are made intelligently with serious intent". No one is saying these companies aren't serious about it, they are saying there are legitimate physics limitations involved here that are either being ignored or the companies are betting on a novel scientific breakthrough.

> your take on investors is naive and largely incorrect - its the musical chairs theory of markets.

Then you clearly have never worked with investors before.

Here’s our point of disagreement: I think smart people have made a bet in this with serious intent after considering all the pitfalls. You think they are either deliberately ignoring it (so ridiculous) or they are betting on a scientific breakthrough.

It’s too comical to even address the “they are ignoring it part” so I’ll ignore it.

I would agree with you that part of their bet might involve hoping for breakthroughs and the investment analysis probably factored it in.

Lots of earlier investments have banked on breakthroughs like this.

Also your opinion on how markets work is naive and unscientific.

> Also your opinion on how markets work is naive and unscientific.

You sure about that?

https://academic.oup.com/rfs/article-abstract/21/1/19/157439...

Because it will inflate their stock valuations? It's like with fusion energy or going to Mars etc., constantly X years away and currently economically unfeasible.

why do you think their stock will inflate?

I'm not sure if you're joking, but "AI datacenter in space" is the kind of phrase that attracts investors, that's straight from Musk's playbook for keeping the stock trading at ridiculous P/Es, especially now that he is planning SpaceX IPO.

why does it attract investors if it is so obvious that it will fail?

it is a ridiculous conspiracy theory you are trying to assert - musk comes up with an absurd idea that captures investor's attention. its not like he wants to make a good product, he just wants to fool investors. not only that, he fools them, gets the money and then puts said money into this venture that obviously won't work. why does he waste his time into a venture that obviously won't work? who knows

You should ask yourself this, not me, you're the one who blindly believes what Musk says. He also said he was creating a new political party in the US, how's that going? Did you believe him when he talked about landing people on Mars in 2018? It’s 2026. How is boring company going? etc. I think you're overinterpreting what I wrote and projecting. I'm telling you how the physics works, and the physics is simple here: unless you change the physics or discover some exotic, cheap materials, this is 100% not economically viable today or in the near future.

So you have no clue why he's doing it? He's putting money in a thing that will obviously fail.

Either you are way way smarter than him, or he's doing this for some other ulterior motive.

You didn't answer my questions. How is The Boring company going? And in this context, you can also ask: "Is he putting money into something that will obviously fail?"

Also, go back and read how many people who were "smarter than him" there nine years ago:

https://news.ycombinator.com/item?id=14223020

Here Bezos, sundar, Jensen all are invested.

On boring: it’s easy to say in hindsight.

You know why I mentioned hydrogen energy earlier? There was a Financial Times article last month titled "Hydrogen dreams meet reality as oil and gas groups abandon projects", which notes that "Almost 60 major low carbon hydrogen projects—including ones backed by BP and ExxonMobil—have been cancelled" because they weren't economically feasible. Space data centers are in the same place today. It's physics. And none of the people you mentioned have invested in this. They may be interested and might research the topic, but that's not the same thing. I've yet to see any plan that explains how they'll replace failed hardware and manage heat while keeping the whole thing economically feasible.

ok so you are smarter than all of them? and if they had put you in charge instead of the phd's, they might have been better off?

You're using an uninteresting appeal to authority argument again.

So let's talk physics. Are you familiar with the radiative heat-balance problem? You can use the Stefan–Boltzmann law to calculate how many radiators you'd need.

Required area: A = P / (eps * sigma * eta * (Tr^4 - Tsink^4))

Where:

A = radiator area [m^2]

P = waste heat to dump [W]

eps = emissivity (0..1)

sigma = 5.670374419e-8 W/m^2/K^4

eta = non ideal factor for view/blockage/etc (0..1)

Tr = radiator temperature [K]

Tsink = effective sink temperature [K] (deep space ~3 K, ~0 for Tr sizing)

Assuming best conditions so deep space, eps~0.9, eta~1:

At Tr=300K: ~413 W/m^2

At Tr=350K: ~766 W/m^2

At Tr=400K: ~1307 W/m^2

So for 10 MW at 350K (basically around 77°C): A ~ 1e7 / 766 ≈ 13,006 m^2 (best case).

And even in the best case scenario it's only 10 MW and we're not counting radiation from the sun or IR from the moon/earth etc. so in real life, it will be even higher.

You can build 10 MW nuclear power plant (microreactor) with the datacenter included on Earth for the same price.

Show me your numbers or lay out a plan for how to make it economically feasible in space.

you are saying you can stop an entire division in google, nvidia, blue origin with this bit of theory?

like all the employees had to do with read this and be like: wow i never saw it that way.

Because "investors" are a large group. Many of them are not involved in the industry and are clueless about tech. Same reason they invest in OpenAI, that hasn't made any money.

Investors, both commercial and individual, often have more money than sense.

Do you think investors as a large group make enormous amounts of money on average?

I didn't say they make it. They have it, like an older person who grew their portfolio over time. They are an example of someone who invests in AI without knowing anything about what it is.

its incorrect - investors and wall street in general _makes_ money on average.

Well it made you invest in them.

Yes because as I said I believe it will work out, so does Google, SpaceX, Blue Origin.

Kodak made a crypto coin. Where did that end up?

> Do you want to run a hypothetical exercise on how many they get right vs wrong? And based on that we can see if this is a "fantasy" or not?

Serious question: If you are so sure that this is a big payday, have you put all your net worth into SpaceX? Seems like a no brainer if you fully believe it.

The reason for this "data centers in space" is the same as the "sustained human colony on Mars". It is all pie in the sky ideas to drive valuation and increase Musk's wealth.

Just a small sampling of previous failed Musk promises: - demonstration drive of full autonomy all the way from LA to New York by the end of 2017 - "autonomous ride hailing in probably half the population of the U.S. by the end of the year" - “thousands” of Optimus humanoid robots working in Tesla factories by the end of 2025." - Tesla semi trucks rollout (Pepsi paid for 100 semis in 2017, and deliveries started in 2022, and now 8 years later they have received half of them.)

I have put as much money as I believe in it (risk adjusted). And same goes for Google, Spacex, Blue Origin and other companies.

This trope

> It is all pie in the sky ideas to drive valuation and increase Musk's wealth.

Really needs to stop. This is based on a naive interpretation of how wealth gets created. Musk has an amazing reputation getting things done and making things that people like. Whether you like him as a person or not, he has done stuff in the past and that's reason enough to believe him now.

>Musk has an amazing reputation getting things done and making things that people like

Are you trolling?

Almost everyone in the industry thinks this way.

SpaceX and Starlink are privately held.

The demand for allocation in any of SpaceX private funding raises is practically unlimited. It’s just not something you can put your money in.

The thing about Elon is that he's got more than enough credibility with betting on big crazy ideas that he's one of a few people that you have to take seriously.

Tesla from LA to NY - https://www.thedrive.com/news/a-tesla-actually-drove-itself-...

Thousand of Optimus Robots...just announced closing a factory to have them focus on robot production - https://www.cnbc.com/2026/01/28/tesla-ending-model-s-x-produ...

SpaceX rewriting the entire economic formula for space launches, accounting for almost 90% of all launches globally last year, becoming a critical piece of the Department of Defense while also launching Starlink globally.

Neuralink let's people control computers with their brain, even playing video games. They're working on an implant to cure blindness right now.

https://neuralink.com/trials/visual-prosthesis/

I get that the man is politically unpopular in some circles, but it's really difficult to bet against him at this point. So far, the biggest criticism has been that it took a little longer than he initially said to deliver...but he did deliver.

Not even a little; doesn’t pass napkin math. It doesn’t solve any problems while adding a litany of new ones: massive radiators for heat rejection, radiation hardening, and enormous launch + repair costs (assuming repairs are even possible). The idea exists to separate investors from their money; the product is the funding round.

I haven't done the actual math and I might be a few orders of magnitude off but shouldn't electrical resistance drop quite significantly in space, too? (Of course there's the other issue that information processing is an inherently dissipative process because entropy etc.)

How would electrical resistance drop in space? If you're thinking "because it's cold" that's actually the biggest issue. The vacuum means you can't dispose of heat easily, so you need giant radiators, which are expensive, heavy, etc.

there's no repair involved. imagine a series of throwaway satellites on an orbit that essentially leaves them close enough together for effective mesh networking, and probably on an orbit that slowly takes them away from earth.

the compute is used for training, not inference. the redundancy and mesh networking means that if any of them die, it is no big deal.

and an orbit that takes them away from earth means they avoid cluttering up earth's orbital field.

It sounds like you're describing Google's proposal, which I believe is at least feasible (though likely uneconomic) unlike, say, Starcloud's. I don't think you are correct about the orbit, though; Google's proposal lists the satellites at 650 km, which would give them approximately 20 years in orbit without boosts. They list estimated life at 5 years given radiation concerns, so they almost certainly would purposely deorbit them earlier.

I've always assumed that the answer to this would be no. However, I also always assumed that a huge space-based internet system would be both expensive and impractical for bandwidth and latency.

Starlink has largely defied those expectations thanks to their approach to optimize launch costs.

It is possible that I'm overlooking some similar fundamental advancement that would make this less impractical than it sounds. I'm still really skeptical.

Wouldn't it be cheaper to buy a container ship, convert it to a data center and connect it to a bunch of floating panels?

Most systems use fresh water for cooling. Salt water can corrode pipes and deposit sediment, but maybe there's a way to use ocean water efficiently.

They also need to be powered and connected to a network, but that seems like an easier problem.

the internal loop can be fresh water which heat exchanges with the ocean water

I wouldn’t overthink the SpaceX / xAI thing. Seems to me it’s a pure financing play to blend two companies owned by the same guy that might look meh on their own to the market but have a compelling narrative about “future growth” together.

All so that the same guy who is already quite rich can continue to run his funny-up money roll-up machine, re-capitalize on a bunch of froth and leave other people holding the bag.

what about Google? It's always the same tired thought ending cliches - companies with "bad" people do obviously "bad" thing to convince idiotic shareholders and prop up the bubble.

i keep seeing this same repeated trope again and again.

Edit: Elon, Sundar, Jensen, Jeff are all interested in this. Even China is.

What conspiracy is going on here to explain it? Why would they all put money into this if it is so obvious to all of you that it is not going to work?

Yep. That was my first thought when I saw the SpaceX & xAI merger news.

https://www.wsj.com/tech/bezos-and-musk-race-to-bring-data-c...

https://www.nytimes.com/2026/01/01/technology/space-data-cen...

Is this satire? I can't even tell anymore. If so, bravo.

Reply on news.ycombinator.com