Jared Vennett (narration):
"In the years that followed, hundreds of bankers and rating agency's executives went to jail. The SEC was completely overhauled, and Congress had no choice but to break up the big banks and regulate the mortgage and derivatives industries."
"Just kidding. Banks took the money the American people gave them, and they used it to pay themselves huge bonuses, and lobby the Congress to kill big reform. And then they blamed immigrants and poor people, and this time even teachers."
Because this move is entirely financial engineering to hide losses just like the roll up of X in to xAI.
None of this has anything to do with business or innovation. Do you not immediately see that? Most of my friends reaction to this news was that this is so obvious it's almost funny (or actually it is funny, since most were laughing as they read the headline).
I'm curious how you could not understand the relevance of the quote unless you were aggressively trying to not understanding it.
Legitimately, did you not immediately conclude it was for financial shenanigans? What did you think? I'm not trying to be shitty, but what else could there be?
This merger smells like a bubble. Servers in space? They don't make enough to cover costs here on earth. Americans will be forced to bail this mess up because we need Falcon 9, Starship one, etc.
I think the argument is that it's messed up that a large debt swap from xAI kept Musk's margin on Twitter from being called by his investors, and now that debt is being absorbed by SpaceX.
I really don't know what you're trying to say. From your comment alone the conclusion I drew was that we should spend more on the industries that makes physical instead of financial products, such as SpaceX.
The sequence of events: Elon doing a leveraged buyout of X, then xAI funding, then debt transfers to X, then the xAI–X stock deal. Now the proposed SpaceX–xAI merger appears to have shifted X’s financial burden from Musk personally toward xAI investors and, potentially, future SpaceX shareholders.
This is speculative, of course, but yeah seems likely.
Who is SpaceX’s biggest customer? And which industry are we being told by any number of governments around the world is the most importantly thing ever and must be subsidized and forced on people?
It is estimated that Starlink is, accounting for 70% - 80% of revenue. Sources: [1] and [2]
NASA is SpaceX's biggest external customer for rocket launch services.
Although NASA is SpaceX’s largest external customer for traditional launch services, the company earns far more revenue from Starlink customers (millions of subscribers). So overall Starlink itself is SpaceX’s biggest revenue generator and de facto largest customer segment.
SpaceX saves its biggest customer money by being the superior, cheaper launch option. The alternative was ULA, which was an extraordinarily expensive monster.
Please highlight the problems you have with how it pertains to this context, how the biggest customer is harmed.
What do you care if its private owners are willing to absorb the mess that is xAI?
> it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power
We currently make around 1 TW of photovoltaic cells per year, globally. The proposal here is to launch that much to space every 9 hours, complete with attached computers, continuously, from the moon.
edit: Also, this would capture a very trivial percentage of the Sun's power. A few trillionths per year.
Context missing. This is in reference to a vision the (distant?) future where the satellites are manufactured in factories on the Moon and sent into space with mass drivers.
Full paragraph quote comes from:
> While launching AI satellites from Earth is the immediate focus, Starship’s capabilities will also enable operations on other worlds. Thanks to advancements like in-space propellant transfer, Starship will be capable of landing massive amounts of cargo on the Moon. Once there, it will be possible to establish a permanent presence for scientific and manufacturing pursuits. Factories on the Moon can take advantage of lunar resources to manufacture satellites and deploy them further into space. By using an electromagnetic mass driver and lunar manufacturing, it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power.
>
We can make ten or hundred times the number of solar cells we make right now, we just don't have a reason to. The technology is fairly ancient unless you want to compete on efficiency, and the raw materials abundant.
What do you think the limiting factor is? I don't see why we can't scale manufacturing of satellites up as far as we want. If we mine out a substantial fraction of the mass of the earth, we can go harvest asteroids or something.
You need to be able to harness enough raw material and energy to build something that can surround the sun. That does not exist in the solar system and we do not yet have the means to travel further out to collect, move, and construct such an incredibly huge structure. It seems like a fantasy.
There are only so many people who can make satellites; there are only so many things to make satellites out of; and there are only so many orbits to put them in. There are only so many reasons why a person might want a satellite. There are only so many ways of placing satellites in orbit and each requires some amount of energy, and we have access to a finite amount of energy over time.
Finally, if we limited ourselves to earth-based raw materials, we would eventually reach a point where the remaining mass of the earth would have less gravitational effect on the satellite fleet than the fleet itself, which would have deleterious effects on the satellite fleet.
Seven reasons are intuitive; I’m sure there are many others.
After a few decades, you need to start replacing all the solar panels.
And the robot army being used to do the construction and resource extraction will likely have a much shorter lifespan. So needs to be self-replicating/repairing/recycling.
Yeah, but besides not having the amount of having the physical amount of material available in the solar system, or the availability of any technology to transfer power generated to a destination where it can serve a meaningful purpose in the foreseeable future, or having the political climate or capital necessary for even initiating such an effort, or not being able to do so without severely kneecapping the habitability of our planet, there are aren't really any meaningful barriers that I can see.
The quoted "1 TW of photovoltaic cells per year, globally" is the peak output, not the average output. They're only about 20% higher peak output in space… well, if you can keep them cool at least.
The 1TW is the rated peak power output. It's essentially the same in space. The thing that changes is the average fraction of this sustained over time (due to day/night/seasons/atmosphere, or the lack of all of the above).
It's still the same 1TW theoretical peak in space, it's just that you can actually use close to that full capacity all the time, whereas on earth you'd need to over-provision substantially and add storage, so 1TW of panels can only drive perhaps a few hundred GW of average load.
The dominant factor is "balance of system" aka soft costs, which are well over 50%.[0]
Orbit gets you the advantage of 1/5th the PV and no large daily smoothing battery, but also no on-site installation cost, no grid interconnect fees, no custom engineering drawings, no environmental permitting fees, no grid of concrete footers, no heavy steel frames to resist wind and snow loads. The "on-site installation" is just the panels unfolding, and during launch they're compact so the support structure can be relatively lightweight.
When you cost building the datacenter alone, it's cheaper on earth. When you cost building the solar + batteries + datacenter, it (can be) cheaper in space, if you build it right and have cheap orbital launch.
The Technology Connections Youtube channel recently did a great video arguing pretty convincingly that the land used to grow corn for cars would be vastly more efficiently used from an energy perspective if we covered it with solar panels.
You just have to remember, most of these people live in high density regions and have little comprehension about how much surface area humanity truly occupies... And that isn't even accounting for offshore constructs.
Realizing the impracticality of it (and that such approaches often collapse under the infeasibility of it) ... wouldn't it be better to... say... cover the Sahara in solar panels instead? That's gotta be cheaper than shipping them into space.
Satellites can adjust attitude so that the panels are always normal to the incident rays for maximum energy capture. And no weather/dust.
You also don't usually use the same exact kind of panels as terrestrial solar farms. Since you are going to space, you spend the extra money to get the highest possible efficiency in terms of W/kg. Terrestrial usually optimizes for W/$ nameplate capacity LCOE, which also includes installation and other costs.
For one or a few-off expensive satellites that are intended to last 10-20 years, then yes. But in this case the satellites will be more disposable and the game plan is to launch tons of them at the lowest cost per satellite and let the sheer numbers take care of reliability concerns.
It is similar to the biological tradeoff of having a few offspring and investing heavily in their safety and growth vs having thousands off offspring and investing nothing in their safety and growth.
Solar modules you can buy for your house usually have quoted power ratings at "max STC" or Standard Testing Conditions, which are based on insolation on Earth's surface.
I think it's because at this scale a significant limit becomes the global production capacity for solar cells, and SpaceX is in the business of cheaper satellites and launch.
> And then there’s that pesky night time and those annoying seasons.
The two options there are cluttering up the dawn dusk polar orbit more or going to high earth orbit so that you stay out of the shadow of the earth... and geostationary orbits are also in rather high demand.
That's pretty much a solved problem. We've had geostationary constellations for TV broadcast at hundreds of megabytes for decades now, and lasers for sat-to-sat comms seems to be making decent progress as well.
> it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power
I grew up on a rural farm in California with a dial-up connection that significantly hampered my ability to participate in the internet as a teenager. I got Starlink installed at my parents' house about five years ago, and it's resulted in me being able to spend considerably more time at home.
Even with their cheapest home plan, we're getting like 100 Mbps down and maybe 20 to 50 up. So it's just not true at all that you would have connections that are a megabit or two per second.
That's not what I'm suggesting. The post says "deep space". If you're going to try to harvest even a tiny percentage of the sun's energy, you're not doing that in Earth's orbit. The comparison is a webcam feed from Mars.
This is quite frankly the largest amount of bollocks Musk has come up with so far. The market is basically a meme market:
Same some futuristic sounding shit -> stonks go up -> repeat
All literally while trying to do really bad things everywhere which are going to end up with bodies piled to the fucking ceiling. And not delivering products. And being a proto-nazi.
This needs to break. Now. No one should be rewarding this shit.
What kind of nonsense is that. SpaceX 2024 revenue barely broke $10B, if that. Launch was probably ~$4B and Starlink probably ~$5B. I'll give them the benefit of the doubt and double those just for shits and giggles and that's still less than $20B and you're claiming >$100B? Horse shit. Nonsense.
A former NASA engineer with a PhD in space electronics who later worked at Google for 10 years wrote an article about why datacenters in space are very technically challenging:
I don't have any specialized knowledge of the physics but I saw an article suggesting the real reason for the push to build them in space is to hedge against political pushback preventing construction on Earth.
I can't find the original article but here is one about datacenter pushback:
But even if political pushback on Earth is the real reason, it still seems datacenters in space are extremely technically challenging/impossible to build.
We don’t even have a habitable structure in space when the ISS falls, there is no world in which space datacenters are a thing in the next 10, I’d argue even 30 years. People really need to ground themselves in reality.
Edit: okay Tiangong - but that is not a data center.
Elon can't legally financially entangle Tesla to SpaceX due to Tesla being a public company, so his hands are tied.
Tesla is clearly benefiting from protectionism and its sales would collapse if BYD were allowed to openly sell in the US. Most people just want affordable, maintainable and reliable cars.
> Elon can't legally financially entangle Tesla to SpaceX due to Tesla being a public company, so his hands are tied.
He absolutely could do it, just like he did when Tesla bought SolarCity. It just isn’t as easy when one of the companies is public than when both are private.
> Elon can't legally financially entangle Tesla to SpaceX
Bill Ackman has proposed taking SpaceX public by merging it with his Pershing Square SPARC Holdings, distributing 0.5 Special Purpose Acquisition Rights (SPARs) to Tesla shareholders for each share held. Each SPAR would be exercisable for two shares of SpaceX, aimed at enabling a 100% common stock capitalization without traditional underwriting fees or dilutive warrants.
With SpaceX IPO set to be one of the biggest of all time, this could have a pretty gnarly financial engineering impact on both companies -- especially if the short interest (direct or through derivatives) remains large.
Why would SpaceX go public? They already have a robust enough private market to give liquidity to all of their employees and shareholders who want it. They can get more private investment.
Going public would add a lot of hassle for little to no gain (and probably a negative of having to reveal their finances).
It has been widely reported for weeks that SpaceX is planning to go public in a few months. The reason is they have big plans to run a vast network of AI servers in orbit and will need to raise a massive amount of funding. xAI merger fits with that plan. I'd assume SpaceX still plans to go public.
Was ignored on HN but here's an article explaining:
it wasn't ignored on HN, there were many articles correctly noting that building data centers in space is a stupid stupid idea because cooling things there is infeasible
AI datacenters are bottlenecked by power, bandwidth, cooling, and maintenance. Ok sure maybe the Sun provides ample power, but if you are in LEO, you still have to deal with Earth's shadow, which means batteries, which means weight. Bandwidth you have via starlink, fine. But cooling in space is not trivial. And maintenance is out, unless they are also planning some kooky docking astromech satellite repair robot ecosystem.
Maybe the Olney's lesions are starting to take their toll.
The shadow thing can be solved by using a sun-synchronous orbit. See for example the TRACE solar observation satellite, which used a dawn/dusk orbit to maintain a constant view of the sun.
The cooling is the bit where I'm lost on, but it will be interesting to see what they pull off. It feels like everyone forgets Elon hires very smart people to work on these problems, it's not all figured out by Elon Musk solely.
He's broken pretty much all the other financial rules.... for example, the amount of blatant self-dealing he gets away with is staggering.
As long as the consequences of his actions continue to increase the paper value for investors, regulations don't really have teeth because there aren't damages. So the snowball gets bigger and the process repeats.
It's "ironic?" considering Tesla launching in China is what created the necessary supply chain to turn BYD into the powerhouse it is today. Tesla's greed will become their own demise.
Tesla cars made in Shanghai are sold in Europe and other places. That is helping them be competitive and they haven't had much price pressure until recently. Just because the Chinese have their own internal competition and deflation which drove their prices down aggressively doesn't mean it was a bad idea to build there. Also the idea the Chinese couldn't figure it out without an American company coming there first to show them is pretty silly.
Tesla Shanghai opened in 2019
BYD made their first hybrid in 2008 and they were a battery company since the 90s
BYD are just affordable and maybe reliable, regarding maintenance their spares are hard to come by and are almost as hard to work with as Tesla and other brands.
I've done plenty of work on my own Tesla. It's not hard to work on at all. Parts are not even very difficult. There are plenty of 3rd party shops (such as one I went to when I needed to replace my windshield.) I really wonder why people continue to think this. It's not 2016 any more.
Tesla body work is extremely expensive. Aluminum, extensive welding instead of fasteners, substantially reduced modularity due to castings, specialized tooling just off the top of my mind.
Starship has a large number of critical milestones coming: Can it land and quickly reuse the upper stage? If not, it can't make refueling flights without building a dozen or two starships. Can it carry the full specified payload? If not, it can't even try to refuel in orbit. If it can't refuel in orbit, it can't go beyond earth orbit. Etc.
Everything has to go right or it will be irrelevant before it works.
> Everything has to go right or it will be irrelevant before it works.
Starship is not all of SpaceX. Saying, maybe because one hates Musk, that SpaceX is going to become irrelevant is wishful thinking.
In 2025 SpaceX launched more rockets into space than the entire world ever sent in a year up to 2022, something crazy like that.
Then out of, what, 14 000 active satellites in space more than half have been launched by SpaceX.
SpaceX is, so far, the biggest space success story of the history of the human race (and GP is right in saying that SpaceX is now a national security matter for the US).
Model S was the most successful EV. If you think cybercab is the vehicle of the future, look at the timeline of the only robo taxi in commerce in the US.
Everything has to go right with that, or cybercab will be irrelevant before it works. Same deal. Same bullshitter.
Model S was successful until Model 3/Y blew it out of the water. Waymo’s timeline is not relevant because they lose money on every car and every deployment. Tesla’s the only financially successful developer of self driving. They can scale it up much faster.In fact, instead of making $5k per car produced, cybercab will net them $50k per car per year.
Haven't you heard? Tesla is pivoting to building humanoid robots instead. They haven't sold a single one, but it toootally warrants retooling their car factories, pinky promise!
Oh c'mon now. Damn model 3 and model S I have driven were considerably lower quality interiors than an ass end Citroen or Fiat. The Model S, a 2023 model the doors didn't even fit properly. And that was all Europe.
As for FSD, nope. Unless you redefine the word reliable.
Edit: I owned a 2018 Model S as well. Literally the worst fucking car I have ever owned or driven.
> BMW actually has a reasonable control surface though, not a grand user interface experiment by some crack heads.
Really? It's one thing to hate Elon Musk, but you're talking about a lot of brilliant engineers who worked on these cars, everything from the components to the software. It's uneeded low blow just because you don't like Elon Musk.
SpaceX is slated to go public some time this year - June IIRC
The biggest selling point /was/ that Musk was being managed there, he wasn't tinkering with SpaceX like Twitter or Tesla, and his foolhardy direction was kept out of the company.
BUT, like Tesla, Musk cannot help himself and is making SpaceX look like a very bad investment - tying his other interests with SpaceX, allegedly using SpaceX money as a "war chest" in his battles.
There is also a danger that investors will see xAI as politically dangerous, which will really hurt SpaceX IPO
They want to go public, but have to sell the hell out of it in the meantime.
I'll bet SpaceX financials aren't as great as some people think. Remember, Elon was the guy who tried to take Tesla private, and talked a lot of smack about how silly it is to be a public company. All of a sudden he wants SpaceX to go public?
Musk has a pattern here - he used Tesla the same way, diverting resources to xAI and treating it as a funding vehicle for other ventures. Once he started doing that, Tesla's financials got murky and harder to trust. Now he's doing it with SpaceX right before the IPO. For investors, that's not 'too big to fail' protection - it's a red flag that the company finances are entangled with his personal empire instead of focused on the core business.
> The biggest selling point /was/ that Musk was being managed there, he wasn't tinkering with SpaceX like Twitter or Tesla, and his foolhardy direction was kept out of the company
The biggest selling point to who? Definitely not wall street
Let’s be honest - this is just a way to prop up Twitter/X. It makes SpaceX shareholders subsidize X, and also American taxpayers who are giving contracts to SpaceX for highly sensitive things. The government should ideally refuse to give SpaceX work unless it unwinds this.
Because Twitter/X is distorting our politics (with ann unbalanced scheme of censorship / amplification / suppression) and destroying the country by mainstreaming far right supremacist politics. Twitter/X does not deserve a single dollar of taxpayer money. If SpaceX is now part of that machine, it doesn’t deserve a single dollar either. I would rather pay more for alternatives and encourage their growth. I also look at any money given to this company as the equivalent of GOP campaign funding, so I feel it should be treated as illegal under the law.
Shouldn't the government be aiming to pay the lowest price for the best goods and services rather than using procurement as a way to promote or suppress certain political opinions?
The government is prevented from doing that by a little thing called the first amendment. "Mainstreaming far right supremacist politics" is just a hyperbolic way of saying he has politics you don't like and is exercising his freedom of the press by promoting it on the media platform he owns. Legally that is no different then the rights that every newspaper and TV station in the country has.
Musk is, indeed, allowed under the 1st to promote whatever he wants to promote. Him being a hypocrite about "free speech absolutism" is not a crime.
However, the current US administration appears to be actively violating the 1st and 5th in a bunch of ways, the 14th that one time, and making threats to wilfully violate the 2nd for people they don't like and the 22nd to get a third term.
I disagree. He would be using taxpayer money to boost his preferred speech. And it is essentially campaign funding for the GOP. It should be treated as such.
> > Boeing is too important to fail as well but it’s been terrible as a shareholder
Your opinion on Boeing being terrible as a shareholder vis-a-vis Tesla would be completely reversed if dividends and capital gains of the 2 companies were to be offered in the form of miles to be flown on Boeing planes and miles on Teslas Uber/Taxi/Autonomous taxis instead of dollars
The absolute overperformance on the stock market that Tesla has enjoyed vis-a-vis Boeing is not rooted in a concrete and tangible quality of life improvement for citizens. Not American citizens, nor global citizens for that matter.
It is my opinion that for all public companies in which it is possible to do so government should mandate payment in kind to all shareholders and board members to prevent the excessive promotional , cult and all around BS aspect of marketing to take over and allow people to profit just by riding off those, and Musk is the GOAT at that.
What do you mean "let it fail?" SpaceX has the most profitable launch system in the world and now operates >50% of all satellites in orbit. They aren't exactly in need of a bailout.
When a company is operating at a scale where you are making orders of magnitude more orbital launches than NASA, operating a constellations of 10,000+ satellites, providing internet access to 10s of millions of people and 1 army, has raised $10s of billions in private markets at valuations in the $100s of billions, then the burden of proof is on you claiming the opposite.
The proof is that they are continuing to launch more mass into orbit than any other entity on the planet - while holding share liquidity events for their employees multiple times a year where they buy back shares. Proof is that they charge a lower cost to orbit than any of their competitors and has done so for years now.
Their revenue from Starlink is slated to be bigger than the entire NASA budget this year.
I am no fan of Musk the man. SpaceX is a strong company and Falcon is a solid vehicle. There is not a lot of competition, and NASA trying to in-source design and supply and construction of a new, reusable LEO rocket would be a complete nightmare.
I root for a competitive rocket market, but SpaceX is at the moment critical.
NASA is still here. Unfortunately, NATA fell victim to enshitification by government contracting. NASA even if it wants to simply cannot today design and launch a rocket. :(
I'm not talking specifically about SpaceX, although historically the cost of their rockets have been much lower than NASA. I'm being much more general. The public sector doesn't have the same incentives that private companies have, whether it's rockets or any other technology. It's sad, but it's the truth.
I wouldn't be too sure. Depends on NASAs mission profiles and a lot of factors. Falcon heavy can bring 26.7t to GTO in expendable mode and only 8t in reusable mode. Reusable cost of Falcon is US$97 million vs US$150 million expendable.
How much does it cost to develop and maintain the reusability? Is it worth the trade-offs in lower tons to orbit due to more weight? Is it worth it adjusting the payload into smaller units, including developing things like refueling in LEO?
Idk, I'm not on the inside doing those calculations...
Twitter (X) owed $1.3B in debt every year in interest since Musk's takeover. This was before re-financing in a higher interest rate environment. The company was losing $200MM+ per year on ~$5B in revenue before the takeover, and there are reports that revenues have decreased by round 50%.
Best case scenario if we accept those numbers is that X makes $3B per year and about half of that goes immediately out the door in debt payments before paying a cent for the entire business to function.
However, if SpaceX acquires X, that ~$1.5B in interest is a fraction of the $8B In profits SpaceX is allegedly generating annually. Further, they can restructure the debt if it's SpaceX's debt, and not owned by X. Investors will be more likely to accept SpaceX shares as collateral than X.
> The company was losing $200MM+ per year on ~$5B in revenue before the takeover, and there are reports that revenues have decreased by round 50%.
X made a profit last year because they cut costs lower than the drop in ad revenue (which is also slowly recovering). The big question is if they will still be profitable in 2026 year without the US election driving big traffic numbers and ads.
He buys twitter at an inflated valuation. Runs it to the ground to a much lower valuation of $9B. [1] Then, his company Xai buys Twitter at a $33B, inflating the valuation up. Then SpaceX merges with Xai for no particular reason, but is expected to IPO at a $1T+ in the upcoming years. [3]
I’m not that smart, but if I were, I would be thinking this is an extended way to move the losses from the Twitter purchase on to the public markets.
It also makes it impossible for Twitter/X to die, as it deserves. It is by far the most toxic mainstream social network. It has an overwhelming amount of far right supremacist content. So bad that it literally resulted in Vivek Ramaswamy, a gubernatorial candidate in Ohio, to quit Twitter/X - nearly 100% of replies to his posts were from far right racists.
Obviously advertisers have not been fans. And it is a dying business. But rather than it dying, Elon has found a clever (and probably illegal) way to make it so that SpaceX, which has national security importance, is going to prop up Twitter/X. Now our taxpayer dollars are paying for this outrageous social network to exist.
I find HN and the tech circles to be one of the main community pillars holding up X. None of my social friends use it anymore, but links absolutely abound here, and it seems like the standard line is to pretend Elon, Grok, all the one button revenge and child porn etc don’t exist. I truly can’t fathom the amount of not thinking about it it would take to keep using the platform.
It was easy to support SpaceX, despite the racist/sexist/authoritarian views of its owner, because he kept that nonsense out of the conversation.
X is not the same. Elon is actively spewing his ultraconservative views on that site.
Now that these are the same company, there's no separation. SpaceX is part of Musk's political mission now. No matter how cool the tech, I cannot morally support this company, and I hope, for the sake of society, it fails.
This announcement, right after the reveal that Elon Musk reached out to Jeffrey Epstein and tried to book a trip to Little St. James so that he could party with "girls", really doesn't bode well.
It's a shame you can't vote these people out, because I loved places like Twitter, and businesses like SpaceX and Tesla, but Elon Musk is a fascist who uses his power and influence to attack some of the most important pillars of our society.
I'm not the best person to make that case as I can only speculate (land cost, permitting, latency, etc). /Shrug
In all the conversations I've seen play out on hacker news about compute in space, what comes up every time is "it's unviable because cooling is so inefficient".
Which got me thinking, what if cooling needs dropped by orders of magnitude? Then I learned about photonic chips and spintronics.
you think we don't have enough space on earth for a few buildings? this seems like a purely western cope. China seems perfectly able to build out large infrastructure projects with a land area smaller than that of the continentenal USA
ISS radiators run on water and ammonia. Think about how much a kg costs to lift to space and you'll see the economics of space data centers fall apart real fast. Plus, if the radiator springs a leak the satellite is scrap.
I don't pretend to understand the thermodynamics of all of this to do an actual calculation, but note that the ISS spends half its time in the shadow of the earth, which these satellites would not do.
Moving electricity long distance is a lot easier than moving coolant long distances, which puts a soft limit on the reasonable size of the solar array of these satellites. But as long as you stay below that and pick a reasonable orbit it's indeed not too bad, you just have to properly plan for it
It does if you don't turn off the heat source every 30 minutes or so. Since the "datacenters" are targeted at sun synchronous orbits they have 24/7 heat issues. And they convert pretty much all collected energy into heat as well (and some data, but that's negligible). Those GPUs are not magically not generating heat.
Wouldn't the panels themselves need cooling too? The ones on earth generate heat while being in the sun.
There are commercial systems that can use open loop cooling (i.e. spray water) to improve efficiency of the panel by keeping the panel at a optimal temp of ~25C and the more expensive closed loop systems with active cooling recovers additional energy from the heat by circulating water like a solar heater in the panel back.
Twenty-five years after the ISS began operations in low Earth orbit, a new generation of advanced solar cells from Spectrolab, twice as efficient as their predecessors, are supplementing the existing arrays to allow the ISS to continue to operate to 2030 and beyond. Eight new arrays, known as iROSAs (ISS Roll-Out Solar Arrays) are being installed on the ISS in orbit.
The new arrays use multi-junction compound semiconductor solar cells from Spectrolab. These cells cost something like 500 times as much per watt as modern silicon solar cells, and they only produce about 50% more power per unit area. On top of that, the materials that Spectrolab cells are made of are inherently rare. Anyone talking about scaling solar to terawatts has to rely on silicon or maybe perovskite materials (but those are still experimental).
5,000 Starship launches to match the solar/heat budget of the 10GW "Stargate" OpenAI datacenter. The Falcon 9 family has achieved over 600 launches.
The ISS power/heat budget is like 240,000 BTU/hr. That’s equivalent to half of an Nvidia GB200 NVL72 rack. So two international space stations per rack. Or about 160,000 international space stations to cool the 10GW “Stargate” datacenter that OpenAI’s building in Abilene. There are 10,000 starlink satellites.
Starship could probably carry 250-300 of the new V2 Mini satellites which are supposed to have a power/heat budget of like 8kW. That's how I got 5,000 Starship launches to match OpenAI’s datacenter.
Weight seems less of an issue than size. 83,000 NVL72’s would weigh 270 million lbs or 20% of the lift capacity of 5000 starship launches. Leaving 80% for the rest of the satellite mass, which seems perhaps reasonable.
Elon's napkin math is definitely off though, by over an order of magnitude. "a million tons per year of satellites generating 100 kW of compute power per ton" The NVL72's use 74kW per ton. But that's just the compute, without including the rest of the fucking satellite (solar panels and radiators). So that estimate is complete garbage.
One note: If you could afford to send up one of your own personal satellites, it would be extremely difficult for the FBI to raid.
you put the radiators and the rest of the satellite within the shade of the solar panels, you can still make the area arbitrarily large
EDIT: people continue downvoting and replying with irrelevant retorts, so I'll add in some calculations
Let's assume
1. cheap 18% efficient solar panels (though much better can be achieved with multijunction and quantum-cutting phosphors)
2. simplistic 1360 W/m^2 sunlight orthogonal to the sun
3. an abstract input Area Ain of solar panels (pretend its a square area: Ain = L ^ 2)
4. The amount of heat generated on the solar panels (100%-18%) * Ain * 1360 W / m ^ 2, the electrical energy being 18% * Ain * 1360 W / m ^ 2. The electrical energy will ultimately be converted to computational results and heat by the satellite compute. So the radiative cooling (only option in space) must dissipate 100% of the incoming solar energy: the 1360 W / m^2 * Ain.
5. Lets make a pyramid with the square solar panel as a base, with the apex pointing away from the sun, we make sure the surface has high emissivity (roughly 1) in thermal infrared. Observe that such a pyramid has all sides in the shade of the sun. But it is low earth orbit so lets assume warm earth is occupying one hemisphere and we have to put thermal IR reflectors on the 2 pyramid sides facing earth, so the other 2 pyramid sides face actual cold space.
6. The area for a square based symmetric pyramid: we have
6.a. The area of the base Ain = L * L.
6.b. The area of the 4 sides 2 * L * sqrt( L ^ 2 / 4 + h ^ 2 )
6.c. The area of just 2 sides having output Area Aout = L * sqrt( L ^ 2 / 4 + h ^ 2 )
7. The 2 radiative sides not seeing the sun and not seeing the earth together have the area in 6.c and must dissipate L ^ 2 * 1360 W / m ^ 2 .
8. Hello Stefan-Boltzmann Law: for emissivity 1 we have the radiant exitance M = sigma * T ^ 4 (units W / m ^ 2 )
9. The total power exited through the 2 thermal radiating sides of the pyramid is then Aout * M
10. Select a desired temperature and solve for h / L (to stay dimensionless and get the ratio of the pyramid height to its base side length), lets run the satellite at 300 K = ~26 deg C just as an example.
11. If you solve this for h / L we get:
h / L = sqrt( ( 1360 W / m ^ 2 / (sigma * T ^ 4 ) ) ^ 2 - 1/4 )
12. Numerically for 300K target temperature we get: h/L = sqrt((1360 / (5.67 * 10^-8 * 300 ^ 4)) ^ 2 - 1/4)
= 2.91870351609271066729
13. So the pyramid height of "horribly poor cooling capability in space" would be a shocking 3 times the side length of the square solar panel array.
As a child I was obsessed with computer technology, and this will resonate with many of you: computer science is the poor man's science, as soon as a computer becomes available in the household, some children autodidactically educate themselves in programming etc. This is HN, a lot of programmers who followed the poor man's science path out of necessity. I had the opportunity to choose something else, I chose physics. No amount of programming and acquiring titles of software "engineer" will be a good substitute for physicists and engineers that actually had courses on the physical sciences, and the mathematics to follow the important historical deductions... It's very hard to explain this to the people who followed the path I had almost taken. And they downvote me because they didn't have the opportunity, courage or stamina to take the path I took, and so they blindly copy paste each others doomscrolled arguments.
Look I'm not an elon fanboy... but when I read people arguing that cooling considerations excludes this future, while I know you can set the temperature arbitrarily low but not below background temperature of the universe 4 K, then I simply explain that obviously the area can be made arbitrarily large, so the temperature can be chosen by the system designer. But hey the HN crowd prefers the layers of libraries and abstractions and made themselves an emulation of an emulation of an emulation of a pre-agreed reality as documented in datasheets and manuals, and is ultimately so removed from reality based communities like physics and physics engineering, that the "democracy" programmers opinions dominate...
It all basically boils down to: in order to dissipate heat, you need something to dissipate heat into, e.g. air, liquid, etc. Even if you liquid cool the GPUs, where is the heat going to go?
On Earth, you can vent the heat into the atmosphere no problem, but in space, there's no atmosphere to vent to, so dissipating heat becomes a very, very difficult problem to solve. You can use radiators to an extent, but again, because no atmosphere, they're orders of magnitude less effective in space. So any kind of cooling array would have to be huge, and you'd also have to find some way to shade them, because you still have to deal with heat and other kinds of radiation coming from the Sun.
That helps with the heat from the sun problem, but not the radiation of heat from the GPUs. Those radiators would need to be unshaded by the solar panels, and would need to be enormous. Cooling stuff in atmosphere is far easier than in vacuum.
Not so. Look at the construction of JWST. One side is "hot", the other side is very, very cold.
I am highly skeptical about data centers in space, but radiators don't need to be unshaded. In fact, they benefit from the shade. This is also being done on the ISS.
Space is not empty. Satellites have to be boosted all the time because of drag. Massive panels would only worsen that. Once you boosters are empty the satellite is toast.
"Satellites have to be boosted all the time because of drag."
On Low Earth Orbits (LEOs), sure, but the traces of atmosphere that cause the drag disappear quite fast with increasing altitude. At 1000 km, you will stay up for decades.
arbitrarily large means like measured in square km. Starcloud is talking about 4km x 4km area of solar panels and radiative cooling. (https://blogs.nvidia.com/blog/starcloud/)
Building this is definitely not trivial and not easy to make arbitrarily large.
I'm not big on this subject, but I understand that heat transfer is difficult in space, because there's little to transfer to. If the solution is just making large radiators, then that means you're sending some big payloads full of radiators. Not to mention all the solar panels needed. I wanna live in sci-fi land too, but I don't see how it makes any sense compared to a terrestrial data center.
If someone has a design out there where this works and you can launch it economically on a rocket today, I wanna see that. And then I wanna compare it to the cost of setting up some data centers on earth (which BTW, you can service in real time, it sounds like these will be one-and-done launches).
You can line the solar panels and radiators facing away from each other, and the radiators would take up less surface area. I think maybe the tricky part would be the weight of water + pipes to move heat from the compute to the radiators.
There's plenty of space in space, but there isn't plenty of space in rocket fairings, nor is there plenty of lift capacity for an unlimited amount of radiators.
It's such bullshit that we've decided this moron and others in his cohort can unilaterally reallocate such vast portions of humanity's labor at their whims.
This is an extremely stupid idea, but because of our shared delusion of capitalism and the idea that wealth accumulation at the top should be effectively limitless, this guy gets to screw around and divert actual human labor towards insane and useless projects like this rather than solving real world problems.
radiators can be made as long as desirable within the shade of the solar panels, hence the designer can pracitically set arbitrarily low temperatures above the background temperature of the universe.
Radiators can shadow each other, so that puts some kind of limit on the size of the individual satellite (which limits the size of training run it can be used for, but I guess the goal for these is mostly inference anyway). More seriously, heat conduction is an issue: If the radiator is too long, heat won't get from its base to its tip fast enough. Using fluid is possible, but adds another system that can fail. If nothing else, increasing the size of the radiator means more mass that needs to be launched into space.
Shading does work; JWST does this. However I don't see how you can make it work for satellite data centers. You would constantly be engaging attitude control as you realigned the panels to keep the radiators in shade. You'd run out of thruster fuel so fast you'd get like a month out of each satellite
Radiators can only be made as long as desirable because there's gravity for the fluid inside to go back down once it condenses. Even seen those copper heat pipes in your PC radiator?
these same comments pop up every time someone brings up satellite data-centers where people just assume the only way of dissipating heat is through convection with the environment.
No, we just "assume" (i.e. know) that radiation in a vacuum is a really bad way of dissipating heat, to the point that we use vacuum as a very effective insulator on earth.
Yes, you can overcome this with enough radiator area. Which costs money, and adds weight and space, which costs more money.
Nobody is saying the idea of data centers in space is impossible. It's obviously very possible. But it doesn't make even the slightest bit of economic sense. Everything gets way, way harder and there's no upside.
Additional radiator area means bigger spacecraft, implies more challenge with attitude control. Lower down you get more drag so you use propellant to keep yourself up, higher up you have more debris and the large area means you need to frequently manoeuvre to avoid collisions. Making things bigger in space is not trivial! You can't just deploy arbitrarily large panels and expect everything to be fine.
The radiators would be lighter compared to the solar panels, and slightly smaller surface area so you can line them back to back
I don't think dissipating heat would be an issue at all. The cost of launch I think is the main bottleneck, but cooling would just be a small overhead on the cost of energy. Not a fundamental problem.
I guess props to scamming Compaq into making a large investment that didn't pan out. He did personally make money so I guess win for him.
>In an effort to woo investors, Elon Musk built a large casing around a standard computer to give the impression that Zip2 was powered by a supercomputer.
>PayPal
Huh? He didn't found Paypal, his company was acquired by Paypal. You might as well give him credit for eBay while you're at it. Paypal released their first digital wallet in 1999. They acquired x.com (and Musk) in 2000. Paypal itself was then acquired by eBay in 2002.
>Tesla
Investor, not founder.
>SpaceX
Yup, props here.
>Grok/xAI
Hasn't made a penny, no signs it had any path to profitability, which is why it was shoved into Space-X to cover his personal losses.
hmm Tesla shipped millions of cars SpaceX launches 90% of space payloads, Starlink is working well. Thats hard to categorize as never delivered on any of his projects
The crucial thing is that Tesla's valuation has the hype projects baked in. The fact that it never delivered self driving or a robotaxi fleet and is now being saved solely by an import ban on Chinese EVs means that any success he had with Tesla is now an illusion.
The projects promised to be life altering for all mankind, they ended up being not even life altering for super rich Americans considering that Teslas are just EVs which without FSD are just regular cars with a different propellent that were made for political purposes and virtue signaling
The EV revolution has always been something almost dystopic : Trillions of dollars spent in order to not have the slightest amount of quality of life improvement, if anything a worse quality of life because you buy an EV that you cannot use 24/7/365 whereas you can an ICE car for much less .
As soon as something kinda elegant and charged with hope as far as collective quality of life improvement is concerned (AI/ChatGPT) the whole green/EV revolution rightfully went out the window
If Musk was this genius you guys make him to be at 50 and with all the capital he burned he should have at least one company that if you disappeared the world would look drastically different, like if you disappeared Microsoft or Apple or Exxon or Aramco the world would come to a screeching halt.
Disappear one of Musk companies and everything would be the same as he's always involved in these sort of aspirational companies which have this great vision always 5 years into the future that never materialize into anything tangible
well Tesla did jump start the EV revolution not life altering but is pretty important. IF SpaceX gets spaceship right that will be a huge leap forward.
Sir, your comment appears to qualify as "moving the goal post". TSLA never delivered a single inexpensive electric vehicle, and just last week abandoned all high-end efforts (S/X/CT discontinued). All TSLA manufactures now are overpriced "meh" transport boxes. Yes, TSLA was early, and now they are far, far behind the competition.
Can we evaluate based on the stated goals, or why does the criteria keep shifting?
In a way, its kind of cool to see how robber barons work in real time in our generation. Its also insanely depressing as they will systematically enshittify and extract as much wealth from society as is possible.
I don't actually think the Robber Barons in the 1920s had people going out of their way to defend them and insist they had special knowledge.
The New Deal happened with massive popular support because people did not like the Barons, and wanted to stop them and actually have a life worth living.
> "The only logical solution therefore is to transport these resource-intensive efforts to a location with vast power and space. I mean, space is called 'space' for a reason. [crying laughing emoji]"
This is all the reasoning provided. It is quite sad how a company I admired so much has become embroiled in financial doohickery.
> The basic math is that launching a million tons per year of satellites generating 100 kW of compute power per ton would add 100 gigawatts of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 1 TW/year from Earth.
> My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space.
This is so obviously false. For one thing, in what fantasy world would the ongoing operational and maintenance needs be 0?
You operate them like Microsoft's submerged data center project: you don't do maintenance, whatever fails fails. You start with enough redundancy in critical components like power and networking and accept that compute resources will slowly decrease as nodes fail
No operational needs is obviously ... simplified. You still need to manage downlink capacity, station keeping, collision avoidance, etc. But for a large constellation the per-satellite cost of that would be pretty small.
How do you make a small fortune? Start with a big one.
The thing being called obvious here is that the maintenance you have to do on earth is vastly cheaper than the overspeccing you need to do in space (otherwise we would overspec on earth). That's before even considering the harsh radiation environment and the incredible cost to put even a single pound into low earth orbit.
If you think the primary source of electricity is solar (which clearly Musk does), then space increases the amount of compute per solar cell by ~5x, and eliminates the relatively large battery required for 24/7 operation. The thermal radiators and radiation effects are manageable.
The basic idea of putting compute in space to avoid inefficient power beaming goes back to NASA in the 60s, but the problem was always the high cost to orbit. Clearly Musk expects Starship will change that.
ISS cooling is 16KW dissipation. So like 16 H200. Now imagine you want to cool 100k instead of 16.
And all this before we talk about radiation, connectivity (good luck with 100gbps rack-to-rack we have on earth), and what have you.
—
Sometimes I think all this space datacenters talk is just a PR to hush those sad folks that happen to live near the (future) datacenter: “don’t worry, it’s temporary”
> ROSA is 20 percent lighter (with a mass of 325 kg (717 lb))[3] and one-fourth the volume of rigid panel arrays with the same performance.
And that’s not the current cutting edge in solar panels either. A company can take more risks with technology choices and iterate faster (get current state-of-the-art solar to be usable in space).
The bet they’re making is on their own engineering progress, like they did with rockets, not on sticking together pieces used on the ISS today.
The idea here is that the economics of launch are changing with Starship such that the "incredible cost" and "overspeccing" of space will become much less relevant. There's a world where, because the cost per kg is so low, a data center satellite's compute payload is just the same hardware you'd put in a terrestrial rack, and the satellite bus itself is mass-produced to not-particularly-challenging specs. And they don't have to last 30 years, just 4-ish, when the computer is ready for retirement anyway.
Will that come to be? I'm skeptical, especially within the next several years. Starship would have to perform perfectly, and a lot of other assumptions hold, to make this make sense. But that's the idea.
How much maintenance do you need? Lets say you have hardware whose useful lifespan due to obsolescence is 5 years, and in 4, the satellite will crash into the atmosphere anyways.
Let's say given component failure rates, you can expect for 20% of the GPUs to fail in that time. I'd say that's acceptable.
If you ramp up the economies of scale to make those things - radiation protection and cost per pound - the calculus changes. It's supposed to synergize with Starship, and immediately take advantage of the reduced cost per pound.
If the cost per pound, power, regulatory burden, networking, and radiation shielding can be gamed out, as well as the thousand other technically difficult and probably expensive problems that can crop up, they have to sum to less than the effective cost of running that same datacenter here on earth. It's interesting that it doesn't play into Jevon's paradox the way it might otherwise - there's a reduction in power consumption planetside, if compute gets moved to space, but no equivalent expansion since the resource isn't transferable.
I think some sort of space junk recycling would be necessary, especially at the terawatt scale being proposed - at some point vaporizing a bunch of arbitrary high temperature chemistry in the upper atmosphere isn't likely to be conducive to human well-being. Copper and aluminum and gold and so on are also probably worth recovering over allowing to be vaporized. With that much infrastructure in space, you start looking at recycling, manufacturing, collection in order to do cost reductions, so maybe part of the intent is to push into off-planet manufacturing and resource logistics?
The whole thing's fascinating - if it works, that's a lot of compute. If it doesn't work, that's a lot of very expensive compute and shooting stars.
Or, just saying, be critical of ideas and think them through, and take in what experts say about it, and determine for yourself what's up. If a bunch of people who usually seem to know what they're talking about think there's a legitimate shot at something you, as a fellow armchair analyst, think is completely impractical, it makes sense to go and see if maybe they know something you don't.
In this case, it's all about Starship ramping up to such a scale that the cost per pound to orbit drops sufficiently for everything else to make sense - from the people who think the numbers can work, that means somewhere between $20 and $80 per pound, currently at $1300-1400 per pound with Falcon 9. Starship at scale would have to enable at least 2 full orders of magnitude decrease in price to make space compute viable.
If Starship realistically gets into the $90/lb or lower range, space compute makes sense; things like shielding and the rest become pragmatic engineering problems that can be solved. If the cost goes above $100 or so, it doesn't matter how the rest of the considerations play out, you're launching at a loss. That still might warrant government, military, and research applications for space based datacenters, especially in developing the practical engineering, but Starship needs to work, and there needs to be a ton of them for the datacenter-in-space idea to work out.
Note how Musk cleverly doesn't claim that not doing maintenance drives down costs.
Nothing in there is a lie, but any substance is at best implied. Yes, 1,000,000 tons/year * 100kW/ton is 100GW. Yes, there would be no maintenance and negligible operational cost. Yes, there is some path to launching 1TW/year (whether that path is realistic isn't mentioned, neither what a realistic timeline would be). And then without providing any rationale Elon states his estimate that the cheapest way to do AI compute will be in space in a couple years. Elon is famously bad at estimating, so we can also assume that this is his honest belief. That makes a chain of obviously true statements (or close to true, in the case of operating costs), but none of them actually tell us that this will be cheap or economically attractive. And all of them are complete non-sequiturs.
Unless I missed something the Microsoft underwater data center was basically a publicity stunt.
Anyone who thinks it makes sense to blast data centers into space has never seen how big and heavy they are, or thought about their immense power consumption, much less the challenge of radiating away that much waste heat into space.
I was listening to a Darknet Diaries episode where Maxie Reynolds seems to make it work: https://subseacloud.com/ I don't know how profitable they are, and I doubt this is scalable enough, but it can work as a business.
Well the thing is that it seemed to have been successful beyond all expectations despite being that? They had fewer failures due to the controlled atmosphere, great cooling that took no extra power, and low latency due to being close to offshore backbones. And I presume you don't really need to pay for the land you're using cause it's not really on land. Can one buy water?
Space is pretty ridicolous, but underwater might genuinely be a good fit in certain areas.
Hot saltwater is the worst substance on earth, excepting, maybe, hydrofluoric acid. You really don't want to cool things with ocean water over an extended period of time. And filtering/purifying it takes vast amounts of power (e.g. reverse osmosis).
An 8 GPU B200 cluster goes for about $500k right now. You'd need to put thousands of those into space to mimic a ground-based data center. And the launch costs are best case around 10x the cost of the cluster itself.
Letting them burn up in the atmosphere every time there's an issue does not sound sustainable.
A Falcon Heavy takes about 63 tons to LEO, at a cost of about $1,500 per kg.
A server with 4x H200s and some RAM and CPU costs about $200k, and weighs about 60kg, with all the cooling gear and thick metal. As is, it would cost $90k to get to LEO, half of the cost of the hardware itself.
I suppose that an orbit-ready server is going to cost more, and weigh less.
The water that serves as the coolant will weigh a lot though, but it can double as a radiation shield, and partly as reaction mass for orbital correction and deorbiting.
Are launch costs really 10x!? Could I get a source for that?
In the back on my head this all seemed astronomically far-fetched, but 5.5 million to get 8 GPUs in space... wild. That isn't even a single TB of VRAM.
Are you maybe factoring in the cost to powering them in space in that 5 million?
You mean you operate them like Microsoft's failed submerged data center project [1]. When pointing at validating past examples you are generally supposed to point at successes.
> By 2024, Project Natick had been inactive for several years, though it was referenced in media as though it was ongoing. That year, Microsoft confirmed that the project was inactive and that it had no servers underwater.
I wouldn't exactly call this a success, for that matter.
But if we’re going down that line of thinking then it’s a poor comparison. I could open a data centre on the ground and use the same principle of zero maintenance, and it would be way cheaper and way more powerful.
This will totally work since we have an unlimited amount of rare earth elements we can just ship off into space never to see again. Infinite raw materials + infinite power equals infinite AI!!!
My understanding was that access to very large body of cold water was a core feature for the project. The water was to be used for cooling relatively efficiently or cheaply.
The thing is: at the end of the day, SpaceX takes the "impossible" and makes it "late".
People are going to Tory Bruno the space datacenters until one day their Claude agent swarm's gonna run in space and they'll be wondering "how did we get here"?
The thing is: at the end of the day, making absolute statements about the inevitability of future success is a fool’s errand.
Musk has a documented history of failing to deliver on promises, timescale or no. So it’s best to engage in some actual critical thinking about the claims he is making.
There's clearly rhetorical hyperbole happening there. But assuming that thermal rejection is good in space, & launch costs continue falling, as earth-based data centers become power/grid-constrained, there is a viable path for space power gen.
The craziest part of those statements is "100 kW per ton." IDK what math he is doing there or future assumptions, but today we can't even sniff at 10 kW per ton. iROSA [1] on the ISS is about 0.150 kW per ton.
Not to be an Elon defender, but can you back up your 0.15/ton? My own searching puts ROSA orders of magnitude higher. Each array is 600kg (0.6t) and puts out 20kw (https://space.skyrocket.de/doc_sdat/irosa-1.htm) which makes 20/0.6 = 33.333 kw/ton
Hey all good. My advice, not that you asked for it, is to put the math in the comment. Even as a footnote. I've found myself backtracking a lot of math comments after I stare at it in the text box for a few seconds.
Maybe I'm just out of the loop, but is solar substantially more efficient in space? I assume the satellites won't orbit in a way that follows the sun. And presumably the arrays of panels they can attach to a satellite don't exceed the size of the panels you could slap on and around a data center (at least without being insanely expensive).
1. solar is very efficient at generating energy, no moving parts, simple physics etc.
2. in space you don't deal with weather or daylight cycle, you can just point your panels at the sun and generate very stable energy, no batteries required
3. environmental factors are simpler, no earthquakes, security, weather. Main problem here is radiation
In theory its a very elegant way to convert energy to compute.
2 is wrong. At a Lagrange point you can do this. Not in low earth orbit - in LEO sunset is every 60 minutes or so, and you spend the next 60 minutes in darkness.
Satellites are heavily reliant on either batteries or being robust to reboots, because they actually do not get stable power - it's much more dynamic (just more predictable too since no weather).
Solar in space is a very different energy source in terms of required infrastructure. You don't need batteries, the efficiency is much higher, cooling scales with surface area (radiative cooling doesn't work as well through an atmosphere vs. vacuum), no weather/day cycles. Its a very elegant idea if someone can get it working.
The panels suffer radiation damage they don't suffer on Earth. If this is e.g. the same altitude orbits as Starlink, then the satellites they're attached to burn up after around tenth of their ground-rated lifetimes. If they're a little higher, then they're in the Van Allen belts and have a much higher radiation dose. If they're a lot higher, the energy cost to launch is way more.
If you could build any of this on the moon, that would be great; right now, I've heard of no detailed plans to do more with moon rock than use it as aggregate for something else, which means everyone is about as far from making either a PV or compute factory out of moon rock as the residents of North Sentinel Island are.
OK, perhaps that's a little unfair, we do actually know what the moon is made of and they don't, but it's a really big research project just to figure out how to make anything there right now, let alone making a factory that could make them cost-competitive with launching from Earth despite the huge cost of launching from Earth.
The US mandates by law that we grow a fuck ton of corn to mix 10% ethanol into gasoline.
If you replaced just those cornfields with solar/wind, they would power the entire USA and a 100% electric vehicle fleet. That includes the fact that they are in the corn belt with less than ideal
sun conditions.
We aren’t even talking about any farmland that produces actual food or necessary goods, just ethanol as a farm subsidy program.
The US is already horrendously bad at land use. There’s plenty of land. There’s plenty of ability to build more grid capacity.
Elon is a pathological liar and it’s crazy that he still gets sanewashed after all he’s done. It’s insanity that he hasn’t been kicked out of leading his companies, and it’s also insanity that he hasn’t been prosecuted by the SEC.
You’ve spent too much life force trying to even understand the liar’s fake logic.
Let’s start right here: there is no such thing as becoming power/grid constrained on earth. If you replaced just the cornfields that the United States uses just to grow corn for ethanol in gasoline just in the corn belt, you could power the entire country with solar+batteries+wind. Easily, and cheaply.
If you don’t even believe that solar+batteries are cheap (they are), fine, choose your choice of power plant. Nuclear works fine.
The truth is, xAI combining with SpaceX is almost certainly corrupt financial engineering. SpaceX as a government contractor and that means Elon’s pal Trump can now siphon money into xAI via the federal government.
This is par for the course for an Elon-associated endeavor but it's been leaking out into the broader tech sector; make ludicrous claims and promises and somehow investors just throw money at you. FSD has been around the corner for over a decade, martian colonization will be here by the end of the decade for the past 20 years and General SuperAI will be here in a few years for the past few years.
Currently, just a cursory google search shows $1500-3000 per kilogram to put something into low earth orbit. Lets take the low bound because of efficiencies of scale. So $1500.
A million tons will cost $1500x1000x1000000= 1,500,000,000,000. That is one and a half TRILLION dollars per year. That is only the lift costs, it does not take into account the cost of manufacturing the actual space data centers. Who is going to pay this?
That launch cost is remarkably cheap to someone that's handled a $1.5million dollar 5U server filled with GPUs and RAM that weighs under 100kg.
Obviously the solar and cooling for the above would both weigh and cost a ton but... It's feels surprisingly close to being within an order of magnitude of current costs when you ballpark it?
Like i don't think it's actually viable, it's just a little shocking that the idea isn't as far out of line as i expected.
Yes, and as we know Starship will be doing regular commercial launches starting in 2020, maybe 2021.
We're getting close to having the time for Starship's delays to be the same as the actual time for the Saturn 5 to go from plans to manned launches (Jan 1962-Dec 1968).
It’s hard to estimate what Starship’s actual costs will be when it isn’t fully operational. I am finding estimates of $100 to $200 per kilogram and even as low as $10 per kilogram.
Let’s say the costs in 5 years do get as low as $15 per kilogram or about 2 orders of magnitude improvement in launch prices. That means a 200-ton payload Starship would cost $3,000 to launch.
Do you honestly believe that? The world’s largest rocket cost a total of $3,000 to launch?
> Let’s say the costs in 5 years do get as low as $15 per kilogram or about 2 orders of magnitude improvement in launch prices. That means a 200-ton payload Starship would cost $3,000 to launch. Do you honestly believe that? The world’s largest rocket cost a total of $3,000 to launch?
You have missed three zeroes in this calculation ;)
15 per kg for a 200-ton payload is about 3 million$. That seems achievable, given that propellant costs are about 1-1.5 million.
"it'll never work" is quite black and white while "failure" is a lot more of a grey area. Will it actually launch? Sure, we've seen it. Will it actually hit the reliability as sold? Will it have as fast of turnaround time to reach launch timing goals? Can it actually launch as much payload as promised? Will the economics actually shake out as intended?
Did the Cybertruck "never work"? Obviously not, they're on the streets. Was it a <$40k truck with >250mi range? No.
Did FSD "never work"? Obviously not, tons of people drive many, many miles without touching the wheel. Does Tesla feel confident in it enough to not require safety operators to follow it on robotaxi trips? No. Does Tesla trust it enough to operate in the Las Vegas Loop? No. Has Tesla managed to get any state to allow it to operate truly autonomously? No.
Look, I hope Starship does work as advertised. Its cool stuff. But I don't see it as a given that it will. And given by the track record of the guy who promised it, it gives even less confidence. I'm sad there's less competition in this space. We have so many billionaires out there and yet so few out there actually willing push envelopes.
One reliable method of pushing envelopes, attracting investment and hiring smart people is to get excited about unrealistic timelines.
The best case is you meed the unrealistic timeline, the average case outcome is you solve the problem but it is delayed several years. And the worst case is it fails and investors lose some money.
If you try to hire people but your message is: we want to reduce the cost of access to space by 20% in thirty years, you are going to get approximately zero competent engineers, and a whole lot of coasters.
And no investors, so you'll be dependent on the government anyway. Depending on the government is great until people you do not agree with or are generally anti science, are in power. I assume this part should not need an example nowadays?
> One reliable method of pushing envelopes, attracting investment and hiring smart people is to get excited about unrealistic timelines.
Its also a good way to shred morale and investor confidence when you're a decade past your timelines or continue to fail on actually delivering on past promises.
I remember reading somewhere that satellites are extra expensive for 2 reasons:
- launch costs are so high that doing exotic bespoke engineering might be worth it if it can shave off a few pounds
- once again because launches are expensive and rare, you cannot afford to make mistakes, so everything has to work perfectly
If you are willing to launch to lower orbits, and your launch vehicle is cheap, you are building in bulk, then you can compromise on engineering and accept a few broken sats
Undergrads afaik even high schoolers have built cubesats out of aluminum extrusions, hobbyist solar panels, and a tape measure as an antenna. These things probably dont do that much, but they are up there and they do work.
It is already more expensive to performance maintenance on SOCs than it is to replace them. Remember, these machines are not for serving a database, there are practically no storage needs (and storage is the component that fails most often.)
Given that, the main challenge is cooling, I assume that will be figured out before yeeting 100 billion $ of computers into space. Plenty of smart people work at these companies.
Do smart people work at Boring Company? Do smart people work on FSD at Tesla? What about the HyperLoop? It is possible for smart people to make technical achievements without the overall project being particularly successful.
I am super intrigued that Elon thinks this. He's got more information about space based compute deployments than any other person in the world, I think -- every Starlink satellite has a linux box in it -- and to my mind the heat problems and hardening problems seem like they would dominate.
But, apparently not, to Elon. I will say there's a MASSIVE cost to getting power infrastructure, land, legal stuff done on terra firma; all that just sort of .. goes away when you're deploying to space, at least if you're deploying to space early and fast.
At any rate, I'm intrigued. For reference, people like Larry Ellison, another hyper-informed genius business man, are investing in and buying small nuclear reactors for power right now on earth, so we are at some sort of inflection point for datacenter demand. I wonder what the first independent service business is going to be in space. Probably satellite clearing / decommissioning.
Sometimes, I wonder how people in the middle ages accepted the whole "Divine Right" of their ruling Kings, while simultaneously suffering under their rule.
> Larry Ellison, another hyper-informed genius business man
"King George, another royal blessed by the divine."
I don't know if you're aware of this, but American markets are hyper competitive. I'd be extremely wary of any instinct to discount the skill level of any top-20 self-made billionaire industrialist, really anywhere in the world, but in the US at least, that skillset is likely heavily skewed toward business excellence.
An off-the-cuff four word description on an Internet forum definitely exceeds the level of worship from a court jester in the 1400s that had to dress up in costume and dance at the command of a king, lest their head get cut off.
Monopoly - so easy everyone can do it. We should give it a go! I would include managing your way into monopoly status and steering clear of being broken up as business skills; no?
The techbro cult is filled to bursting with greedy, narcissistic people who are wholly willing to ignore evil because they expect to be the next dispensers of said evil.
That's the most measured Elon critique I've read today. :)
I've been told by SpaceX folk that Elon's job is to keep a 20 year view in the future and essentially get folks to work backward from that.
I think I might kind of be sold on data-clusters in space in 20 years.
I can understand if I had lift that cost 1/10 what everyone else in the world paid for it, I'd be even more sold on them.
That said, this newfound enthusiasm of his certainly makes a commercially reasonable path forward to turn xAI stock into spacex stock. Elon takes care of his investors, generally speaking.
SpaceX supposedly mostly runs non-rad-hard parts, the ostensible reason being because its more cost effective to double or triple up than buy specialty equipment. Do you have a source for this?
You are conflating with their rockets. They don’t go space grade in their rockets because they don’t need to. They’re not up there for very long, and the avionics can be shielded behind large rocket stages.
> I will say there's a MASSIVE cost to getting power infrastructure, land, legal stuff done on terra firma; all that just sort of .. goes away when you're deploying to space, at least if you're deploying to space early and fast.
You need both power infrastructure and structures to build within for deploying in space too. And you have to build them and then put it all into space.
Cost per square foot of land is not that high basically anywhere you could build a datacentre to offset that.
Well some stuff you either don't need, or just can't have so you do something different - for instance, transformers to convert grid power - no grid - no transformers. Those are like a 36 month wait list in the US right now. And solar is something like 2x as efficient in space.
I agree those don't seem immediately to be huge wins to me; not dealing with local politics might be a big one, though. Depending on location. There's a lot of red tape in the world.
$/sq foot or meter belies the cost of dealing with every regulatory agency that has claim on that area. There's no environmental commission you've got to pay off if your satellite starts leaking noxious chemicals all over the place, the same way you'd have to if you spilled something at NUMMI in Fremont, California.
> He's got more information about space based compute deployments than any other person in the world...
He also had more information about self-driving progress than any other person in the world - yet he was wrong with his predictions every year for last 10 years.
Elon doesn't think this. Elon says whatever bullshit thing comes into his head without regard for technical, economic or physical plausibility. As long as it raises the stock price!
The market has had almost a hundred years of being well-regulated, so when a sociopath lies through their teeth, we're inclined to give them the benefit of the doubt. But in the last few years, that regulation has been worn down to nothing, and the result is and was entirely predictable: fraud.
One of the biggest but most pointless questions I have about our current moment in history is whether the people in power actually believe the stuff they say or are lying. Ultimately I don't think the answer really matters, their actions are their actions, but there is just so much that is said by people like Musk that strains credulity to the point that it indicates either they're total idiots or they think the rest of us are total idiots and I'm genuinely curious which of those is more true.
We’re at a point where propaganda is so much more powerful than reality that the people in power literally can’t tell the difference. When your source of ethics is the stock price, little details like physical impossibility stop seeming relevant.
Here is my main question: Musk is on record as being concerned about runaway "evil AI." I used to write that off as sci-fi thinking. For one thing, just unplug it.
So, let's accept that Musk's concern of evil runaway AI is a real problem. In that case, is there anything more concerning than a distributed solar powered orbital platform for AI inference?
Aside from anything about Elon Musk, here’s an interesting video response to the “just unplug it” argument on the Compuerphile channel: https://youtu.be/3TYT1QfdfsM
Ha, I figured that might be the video prior to clicking it. I am a long time fan.
Agreed, when I wrote "just unplug it," this counterargument was present in my mind, but nobody likes a wall of text.
However, my original point was that a distributed solar powered orbital inference platform is even worse! Think about how hard it would be to practically take out Starlink... it's really hard.
Now.. >1M nodes of a neural net in the sky? Why would someone who lives as a god, the richest man in the world, the only person capable of doing this thanks to his control of SpaceX... do the literal worst thing possible?
tone: I don't really understand orbital mechanics, but I do understand geopolitics a bit.
1. China is very concerned about Starlink-like constellations. They want their own, but mostly they want to be able to destroy competitors. That is really hard.
2. Many countries have single ASAT capabilities. Where one projectile can hit one satellite. However, this is basically shoot a bullet, with a bullet, on different trajectories.
3. > Sure, it'd take orbital launch capabilities to lift ... how many bags of metal scrap and explosives?
If I understand orbital mechanics... those clouds of chaff would need to oppose the same orbit, otherwise it is a gentle approach. In the non-aligned orbit, it's another bullet hitting a bullet scenarios as in 2, but with a birdshot shotgun.
My entire point is that constellations in LEO take hundreds of Falcon 9's worth of mass to orbit and delta-v to destroy them, as in-orbit grenades which approach gently. This IS REALLY HARD, as far as mass to orbit, all at once! If you blow up some group of Starlink, that chaff cloud will just keep in orbit on the same axis. It will not keep blowing up other Starlinks.
The gentle grenade approach was possibly tested by the CCP here:
> tone: I don't really understand orbital mechanics, but I do understand geopolitics a bit.
Thanks for the clarification, I guess that explains this (from you):
> Think about how hard it would be to practically take out Starlink.
and this:
> My entire point is that constellations in GEO
which you've now corrected.
Moving on:
> My entire point is that constellations in LEO take hundreds of Falcon 9's worth of mass to orbit and delta-v to destroy them, as in-orbit grenades which approach gently. This IS REALLY HARD
So let's not do that .. how hard is it to render the entire LEO zone a shit show with contra wise clouds of frag that cause cascading failures?
Forget the geopolitics of China et al. .. LEO launch capabilities are spreading about the globe, it's not just major world powers that pose a threat here.
Ok... so, let's reset, please. I bet that we have very similar intentions, and yet on internet forums, we have perfected the art of users speaking past each other.
Just to get on the same page here. My arugument is that prior to Elon Musk, the only human capable of launching >1M distributed solar powered inference nodes, if one accepts runaway AGI/ASI as a threat... prior to that we had a few hundred terrestrial AI inference mega-data centers. Most of them had easily disrupted power supplies by one dude with a Sawzall.
Now, we are moving to a paradigm where the power supply is the sun, the orbital plane gives the nodes power 24/7, and the dude with the Sawzall needs to buy >10,000x (not sure of the the multiple here) the Sawzalls, and also give them escape velocity.
Can we not agree that this is a much more difficult problem to "just unplug it," than it was when the potentially troublesome inference was terrestrial?
To address your grade-A snark upthread which provided a chuckle, you may indeed have a better grasp of orbital mechanics than I do; I've not worked in the field professionally since I dealt with raw MODIS streams direct from ground stations.
And, moving on in good faith; yes, things can be difficult but not insurmountable or impossible.
Let's call rendering LEO unusable a real actual risk going orward .. there are papers on this.
My attitudes are shaped by having grown up with literal Z-force / SASR types; my father (still alive) had a lifelong friendship with Major General Philip Michael Jeffery, I took theodolite lessons from Len Beadell, diving lessons from Jack Wong Sue, and my aunts flew spitfires, jets, and heavy bombers in WWII.
There are many people in this world who, if asked, would regard taking out a LEO constellation as an interesting challenge.
What, creating a huge patchwork of self sufficient AIs, forming their own sky based net, seems bad to you, considering the whole torment nexus/Sky Net connotations? It's not like he's planning to attach it to his giant humanoid robot program. Oh. Ohhhhh. Oh no.
Any estimate by Elon musk, you need to add or substract a zero to/from the end. Here, I'll fix it for you.
> The basic math is that launching a 100,000 tons per year of satellites generating 10 kW of compute power per ton would add 1 gigawatt of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 0.01 TW/year from Earth.
> My estimate is that within 20 to 30 years, the lowest cost way to generate AI compute will be in space.
But more importantly, there is no heat dissipation in space. There’s no atmosphere to cool you, no water you can put heat into. Just an empty void. You can radiate a little, but the sun alone is enough to cook you, without you having a rack of GPUs inside your satellite.
It’s completely delusional to think you could operate a data centre in a void with nowhere to put the heat.
Never mind operational and maintenance costs. In what fantasy world is it cheaper to put a computer in orbit than in a building on the ground? I don't care how reusable and maintenance-free Starship gets, there's no way even absurdly cheap launches are cheaper than a building.
The whole thing makes no sense. What's the advantage of putting AI compute in space? What's even one advantage? There are none. Cooling is harder. Power is harder. Radiation is worse. Maintenance is impossible.
The only reason you'd ever put anything in orbit, aside from rare cases where you need zero-gee, is because you need it to be high up for some reason. Maybe you need it to be above the atmosphere (telescopes), or maybe you need a wide view of the earth (communications satellites), but it's all about the position, and you put up with a lot of downsides for it.
I feel like either I'm taking crazy pills, or all these people talking about AI in space are taking crazy pills. And I don't think it's me.
The most generous interpretation is that the "AI in space" nonsense is a cover for putting limited AI in space for StarShield (military version of StarLink), which is essentially the "Golden Dome".
It might be possible to scam the Pentagon with some talk about AI and killer satellites that take down ICBMs.
Only if that bitflip happens somewhere in your actual data, vs. some GPU pipeline register that then locks up the entire system until a power cycle. Or causes a wrong address to be fetched. Or causes other nasty silent errors. Or...
Try doing fault injection on a chip some time. You'll see it's significantly easier to cause a crash / reset / hang than to just flip data bits.
'rad-triggered bit flips don't matter with AI' is a lie spoken by people who have obviously never done any digital design in their life.
You do realize that “space-grade” involves process changes that intrinsically incur orders of magnitude efficiency losses? Larger process sizes, worse performing materials. It’s not just a design thing you can throw money at.
The problem is not solved and the techniques they use to deal with it run directly contrary to maximizing compute, because that's not historically something they have remotely cared about.
Elon Musk’s AI startup, xAI (x.ai = grok), officially acquired X Corp. (the parent company of the social media platform X or x.com, formerly Twitter/x.com) in an all-stock deal. Both now operate under a unified holding entity, frequently referred to in corporate filings as X.AI Holdings Corp. (or simply xAI). Now SpaceX has moved to acquire or merge with xAI. This effectively brings the social media data from x.com, the AI development of x.ai, and the satellite infrastructure of Starlink/SpaceX under one "super-conglomerate" roof.
1. What in the circular funding? This feels more like a financing scheme founding it under X/Twitter and then spinning it over to SpaceX. I suspect some debt is disappearing or taxes aren't getting assessed because of this move.
2. The only thing harder than harnessing "a millionth of the sun's power" on Earth would be launching enough material into space to do the same thing. And that's not even a reason for SpaceX to own an AI company, at least not at this point. The current AI isn't going to help with the engineering to do that. Right now hiring 20-somethings fresh out of college is way cheaper and SpaceX has been very successful with that.
quick edit: dang, I even got point 1 backwards. xAI owns X/Twitter, and that means that SpaceX now owns X/Twitter as well as an AI company. Super suspicious that SpaceX could actually think that buying the social media part (a significant portion of xAI's value) would be worth it.
Just a neat bit of financial engineering. You can tell because Elon picked SpaceX instead of Tesla – which would have actually made sense at some level (Optimus Robots + AI). But Tesla is public and so he'd need to follow laws and reporting requirements.
You can tell it's just financial engineering because in the entire press release xAI is only mentioned in the first two sentences. Everything after that is Elon talking about space data centers to distract from the actual topic. Which seems to be working
Tesla is still very profitable, as is SpaceX I assume. Twitter/X has been a $44 billion dollar failure, and xAI is a vanity project so Musk can go around saying he is a player in the AI space. Investors in both X and xAI need to be bailed out, hence this announcement.
In the world where it makes $8-10B in profit on $90-95 billion in revenue every year. Whatever price investors choose to trade the stock at is irrelevant to those numbers.
The $44B Twitter/X buyout was not a failure. For example Fidelity has its $19M investment in the buyout - now xAI common shares - marked at $62M (up over 3X) as of 12/31/25. It was certainly valued even higher on 1/31/26 after xAI had an oversubcribed fund raise in January. All before this merger announcement.
The fact that it had to be successively bailed out by xAI (which itself was funded by Tesla) and now SpaceX shareholders is exactly what makes the acquisition a failure.
He spent other people's money (or maybe even imaginary money) he couldn't have used for himself (since selling off major stakes in your company is a big nono)
Do you genuinely not think that "Elon" (xAI) is player in the AI space?
You don't have to think they have the best models of course, but they are clearly a very significant, and some might argue, leading player in the AI race.
xAI’s models are really not pioneering at all. They weren’t the first to do MoE. Not the first to do open weighting, not the first to have memory or multi-modal vision.
So no, I wouldn’t say Elon is a major player in the AI space. People use his models because they are cheap and are willing to undress people’s photos.
saying they aren't pioneering is very different than saying they aren't a major player in the space. There're only like 5-7 players with a foundational model that they can serve at scale. xAI is one of them
I guess the difference is Tesla is a public company, so requires more paperwork. SpaceX isn't public yet, but will be soon, meaning it will have a cash infusion.
I suspect SpaceX will acquire Tesla at some point. It’s the most profitable of these companies. So basically SpaceX employees and shareholders are covering up for the failing Tesla business and the already-failed xAI business.
Let’s not forget, xAI is the parent of Twitter/X (the social network). So now, taxpayers are paying to keep Twitter/X alive. After all, it is taxpayer money going to the contracts the government gives SpaceX for launches. Nice way to subsidize what is effectively a one sided campaign machine for the GOP and far right.
> I suspect SpaceX will acquire Tesla at some point.
I think that is also likely, unless Tesla can stage a major turnaround, it is going to be beaten by Chinese competitors nearly everywhere that they are allowed (which is everywhere but the USA.)
I get what you're saying, but that taxpayer money is paying for the launch services at a very competitive rate (possibly the cheapest of all available options), not a subsidy scheme.
This was my immediate thought as well. A great time to ask yourself — why am I literally paying for any of this? At best I literally don't use any of these services, at worst they are actively used against me.
I don't see the demand for space being there, OSS is driving costs down and there are still plenty of hardware and algorithmic optimizations we haven't deployed yet.
I am concerned, and haven’t seen anyone else point this out yet, that Musk will move Grok’s CSAM generation capabilities to space to be beyond the reach of terrestrial policing. Does this create some sort of legal loophole here so Musk can do this with impunity?
No. For one, things in space are under the jurisdiction of the country they were launched from. For another, it's people that do crimes, not satellites or LLMs, and the people involved in making CSAM are all on Earth.
AFAIK he can do whatever he wants in space, but CSAM is still illegal to view or even download in most (all?) countries of the world. So unless the degenerates also move out into space (which I'm sure they're eager to do), it wouldn't really ease the legal situation here on Earth.
Maybe if Elon launched himself and the dev team into orbit and didn't use any ground stations and just Starlink terminals he could start getting into legal loopholes.
The reason is probably that Tesla is falling behind on EVs, or at least feels like they've juiced all they can from them at the moment, but advanced robotics is still on the upswing and probably is far from reaching its full potential. They have enough money that moonshots like these probably seem irrelevant at their scale.
As for the space datacenter idea, I think this is just a case of extreme marketing that Musk's ventures are so accustomed to. Making huge promises to pump their stocks while the US government looks the other way. When time comes for them to deliver on their promises, they've already invented ten more outrageous ideas to make you forget about what they promised earlier. Hyperloop as a viable mode of transportation, tunnel networks for Teslas, SpaceX vehicles as a mode of transport, X as the new 'everything app', insane timelines for a Musk-led human mission to Mars. They've done it all.
Send up a spacecraft with back-to-back / equal area solar panels and radiators (have to reject heat backwards, can't reject it to your neighboring sphere elements!). Push your chip temp as much as possible (90C? 100C?). Find a favorable choice of vapor for a heat pump / Organic Rankine Cycle (possibly dual-loop) to boost the temp to 150C for the radiator. Cool the chip with vapor 20C below its running temp. 20-40% of the solar power goes to run the pumps, leaving 60-80% for the workload (a resistor with extra steps).
There are a lot of degrees of freedom to optimize something like this.
Presumably the cooling problem gets hand waved away as a technical detail, and the real selling point is data centers that aren't subject to any regional governments laws.
Cooling a datacenter in space isn't really any harder than cooling a starlink in space, the ratio of solar panels to radiating area will have to be about the same. There is nothing uniquely heat-producing about GPUs, ultimately almost all energy collected by a satellite's solar panels ends up as heat in the satellite.
IMO the big problem is the lack of maintainability.
> Cooling a datacenter in space isn't really any harder than cooling a starlink in space
A watt is a watt and cooling isn't any different just because some heat came from a GPU. But a GPU cluster will consume order of magnitudes more electricity, and will require a proportionally larger surface area to radiate heat compared to a starlink satellite.
Best estimate I can find is that a single starlink satellite uses ~5KW of power and has a radiator of a few square meters.
Power usage for 1000 B200's would be in the ballpark of 1000kW. That's around 1000 square meters of radiators.
Then the heat needs to be dispersed evenly across the radiators, which means a lot of heat pipes.
Cooling GPU's in space will be anything but easy and almost certainly won't be cost competitive with ground-based data centers.
Sure, but cooling a starlink in space is a lot more difficult than cooling a starlink on earth would be. And unlike starlink which absolutely must be in space in order to function, data centers work just fine on the ground.
You can just use the cheap solar panels that were gonna be launched into space (expensive) and not launch them into space (not expensive) and plug them into some batteries (still, cheaper than a rocket launch)
According to Gemini, Earth datacenters cost $7m per MW at the low end (without compute) and solar panel power plants cost $0.5-1.5m per MW, giving $7.5-8.5m per MW overall.
Starlink V2 mini satellites are around 10kW and costs $1-1.5m to launch, for a cost of $100-150m per MW.
So if Gemini is right it seems a datacenter made of Starlinks costs 10-20x more and has a limited lifetime, i.e. it seems unprofitable right now.
In general it seems unlikely to be profitable until there is no more space for solar panels on Earth.
I think that it's not just about the ratio. To me the difference is that Starlink sattelites are fixed-scope, miniature satellites that perform a limited range of tasks. When you talk about GPUs, though, your goal is maximizing the amount of compute you send up. Which means you need to push as many of these GPUs up there as possible, to the extent where you'd need huge megastructures with solar panels and radiators that would probably start pushing the limits of what in-space construction can do. Sure, the ratio would be the same, but what about the scale?
And you also need it to make sense not just from a maintenance standpoint, but from a financial one. In what world would launching what's equivalent to huge facilities that work perfectly fine on the ground make sense? What's the point? If we had a space elevator and nearly free space deployment, then yeah maybe, but how does this plan square with our current reality?
Oh, and don't forget about getting some good shielding for all those precise, cutting-edge processors.
Assuming you can stay out of the way of other satellites I'd guess you think about density in a different way to building on Earth. From a brief look at the ISS thermal system it would seem the biggest challenge would be getting enough coolant and pumping equipment in orbit for a significant wattage of compute.
Because the latencies required for modern AI training are extremely restrictive. A light-nanosecond is famously a foot, and the critical distances have to be kept in that range.
And a single cluster today would already require more solar & cooling capacity than all starlink satellites combined.
Because that brings in the whole distributed computing mess. No matter how instantaneous the actual link is, you still have to deal with the problems of which satellites can see one another, how many simultaneous links can exist per satellite, the max throughput, the need for better error correction and all sorts of other things that will drastically slow the system down in the best case. Unlike something like Starlink, with GPUs you have to be ready that everyone may need to talk to everyone else at the same time while maintaining insane throughput. If you want to send GPUs up one by one, get ready to also equip each satellite with a fixed mass of everything required to transmit and receive so much data, redundant structural/power/compute mass, individual shielding and much more. All the wasted mass you have to launch with individual satellites makes the already nonsensical pricing even worse. It just makes no sense when you can build a warehouse on the ground, fill it with shoulder-to-shoulder servers that communicate in a simple, sane and well-known way and can be repaired on the spot. What's the point?
Remember MoviePass, and how they were losing gobs of money by letting people see unlimited movies for $20/month?
It was so obviously stupid that a bunch of people went, "well, this so clearly can't work that they must have a secret plan to make money, we'll invest on that promise", and then it turned out there was no secret plan, it was as stupid as it looked and it went bankrupt.
The "datacenters in space" thing is a similar play: it's so obviously dumb that a bunch of smart people have tricked themselves into thinking "wow, SpaceX must have actually figured a way it can work!"; SpaceX has not and it is in fact exactly as stupid as it looks.
But it won't end the same as MoviePass until Elon dies; he will keep moving things around, propping up failures with VC, IPO, federal/state (taxpayer) and profit making business money.
My guess would be just a regular radiator and cooling system like a liquid pump. The only obstacle should be the vacuum. That said I don't have any hopes Elon has any understanding of any of it.
There are regularly comments on HN from people who say they pay for Musk's various products, and I am always downvoted into oblivion for suggesting that that is ethically problematic.
There's obviously quite a lot of autocratist illiberals in tech.
I can't tell how many layers of sarcasm are here, but I just want to highlight that aktshually cooling in space is quite difficult because there is no convection, so the only cooling option is radiative. Which gets a bit hard when the satellite gets blasted by the sun.
The ISS doesn't have problems staying warm, it has problems cooling off.
Everything I've heard from Musk in the past decade has been against my will and has made me dumber. (no I do not care to verify or know whether the above is true)
Edit: ah fuck ya got me "the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe" what the cultish bullshit is this. In a just world investors would be fleeing in droves from this cuckoo behavior (I know xAI & SpaceX are private)
A large telecommunications satellite operates at about 15kW. A Blackwell GPU consumes 1kW so you would be at 15 Blackwells per satellite. The cooling surface needs to scale linearly so there is little return to scale.
tbh you could just combine them with starlink sats. didn't they just apply for (and get?) a license for 1 million sats? Stick a single racks worth of gpu power on those and hey presto you've just got yourself the largest ai cluster in the world by far.
But the "fraud" here is being done mostly to VC investors with deep pockets and lawyers, at least until he tries to take this entity public. And I can't imagine them just taking this lying down, but then again maybe they realize that offloading this steaming pile on public market investors is the best way out. But even then... SpaceX seemed like it was quite viable on its own, the investors there are the real losers here.
> Starship will deliver millions of tons to orbit and beyond per year
Excuse my naive physics, but is there a point at which if you take enough mass off of earth and launch it into space, it would have a measurable effect on earth's orbit? (Or if the mass is still tethered to earth via gravity, is there no net effect?)
Earth weighs about 5972200000000000000000000 kg. They are claiming to plan to launch ~5000000000 kg / year. That's 8x10^-14 %. You'd need some pretty accurate instruments to tell the difference.
Related: NASA chief suggests SpaceX may be booted from moon mission [1]. Blue Origin could snatch SpaceX's Starship lander contract. This looks increasingly a good idea.
Sean Duffy is no longer acting administrator of NASA. This proposal was apparently part of a bid to get the support of a coalition of old-space companies and new-space non-SpaceX companies. As part of that strategy he apparently leaked Isaacman's Project Athena document and was backgrounding that he was a SpaceX plant.
But, Isaacman is administrator now, and whatever you think about Isaacman and his relationship to SpaceX, I don't think there's much merit in thinking one of Duffy's half thought out plans is likely to be carried out.
Sadly this seems correct. When Trump was re-elected Elon Musk pushed for Jared Isaacman to be appointed as NASA administrator. When the pick went another way, it led to some real friction between Musk and Trump. Now, with Isaacman finally at the helm of NASA, it looks like Musk’s influence over the agency has come full circle.
The original Moon mission was masterminded by a literal card-carrying ex-member of the Nazi party (Wernher von Braun) and the American public back then didn't seem to mind.
The really skeptical take here is that eventually all of Musk's companies merge, or at least the biggest ones, for juicing that market value to get that $1T payout. Looking at Tesla.
Seems like a way to put a lot of junk in space. If thats in earth orbit it will lead to a lot of junk falling from the sky in 10 years. If it all burns up that will be a lot of nasty shit in the atmosphere - millions of tons!
Kind of a bad look - but I can't precisely say why. Maybe he thinks he can raise more capital this way than he could for each company separately? Especially raising more money for X might be quite hard - they seem to be quite a bit behind on the revenue side compared to OpenAI / Anthropic. With both companies merged he might just find enough retail investors willing to buy at sky high valuations.
Doesn't the idea of Orbital Datacenters imply that the constraining resource right now is physical space, and not compute, electricity, etc?
Did we suddenly solve the electricity problem, or the compute problem?
As far as im aware there are still plenty of datacenters being planned and built right now.
One thing to keep in mind. xAI and SpaceX both have contracts with the DoD. So it makes sense he moved it there rather than Tesla. Not sure I buy the needing AI for doing more in space or if this is to save sinking ship, but if one of his two big companies needed to buy it to keep it afloat it makes sense it was SpaceX and not Tesla.
I'm wondering if SpaceX's going public will be delayed. If not we'll see the first test of the public's appetite for what the AI companies' balance sheets look like
Whenever computer chips go into space, they have to be hardened against radiation, because there is no atmosphere to protect them. Otherwise you get random bit flips.
This process takes a while, which is partly why all the computers in space seem out of date. Because they are.
No one is going to want to use chips that are a many years out of date or subject to random bit flips.
(Although now it got me thinking, do random bit flips matter when training a trillion parameter model?)
I know that per HN's guidelines we're supposed to be "kind and curious", and "reply to the argument instead of calling names". But with some texts, engaging with individual arguments loses sight of the more important bigger picture. So while unkind, the most "thoughtful and substantive" thing I think can be said about this text is:
They must have linked the wrong press release /s. I would have expected a press release about SpaceX acquiring xAI to talk about why they did that. Or at least mention xAI beyond the first paragraph. This is just Elon talking about space data centers
Didn’t Elon say that orbital solar collection was a stupid idea due to energy loss in transmission? Using AI as an almost proof-of-work shows that it may potentially be more complex problem than previously thought. If we threw Bitcoin miners up to those satellites you could literally beam money down.
Musk is moving value out of public hands and into his own. He overpaid $44B for Twitter, then rebranded it as an AI asset by folding X into xAI. He pushed Tesla to invest $2B of shareholder money into xAI despite shareholders voting no. Five days later, SpaceX acquired xAI, effectively turning Tesla’s cash into equity in a private company Musk owns far more of. Musk controlled every step, there was no real arm’s-length process, and he almost certainly knew the outcome in advance. Musk and his private investors get control, inflated valuations, and IPO upside. Tesla shareholders supply the cash, take the risk, and lose leverage.
Just checking (genuine question) there wouldn't be a sneaky way to weaponize a million satellites in orbit around the Earth, would there? I can't imagine it wouldn't have ever been looked into.
This makes me genuinely sad. SpaceX was the one thing of his that Elon has largely avoided screwing up. Imho, this is in large part due to Gwynne Shotwell. She seems to have the personality (not to mention, personal wealth) to kick Elon in the head when he tries to mess things up.
What’s happening now is nothing more than a transparent effort to couple the AI hype-wagon to SpaceX in order to drive the valuation higher in the minds of investors who still think that LLMs will completely transform society.
I’ll be thrilled if the rocket folks can avoid being distracted by this nonsense, but I’m not optimistic.
I’ve been following SpaceX since something like the 2nd Falcon 1 launch and this is the worst thing I’ve seen happen. Sad times.
Let's call it for what it is, a payday for Elon. Paper billionaires have figured out they cannot cash out with out tanking their paper, so now you have these circular deals to extract as much as possible. If we had a functioning government they would step in and put an immediate stop to this on national security grounds.
I suppose one of the ADR’s read something like “…who cares about bitflips, man. Isn’t AI all about probability?”
Knowing the insane level of hardening that goes into putting microcontrollers into space, how to the expect to use some 3nm process chip to stand a chance?
A trend at the moment is to just hope for the best in cubesats and other small satellites in LEO. If you’re below the radiation belt it’s apparently tenable. I worked somewhere designing satellite hardware for LEO and we simply opted to use consumer ARM hardware with a special OS with core level redundancy / consensus to manage bit flips. Obviously some problems will present for AI there… but there are arguably bigger problems with AI data centres like the fact that they offer almost no benefit with respects to the costs of putting and maintaining stuff in space!
I have so many conflicting thoughts that I cannot properly articulate yet. I can say though, this is not going to end well for most, it is clumsily premeditated and starting to feel like dude is just trying to be a Neal Stephenson character.
well, Musk has been overpromising and under delivering for a decade (or more?), so it seems pretty clear this too is shithousery, albeit possibly ambitious.
I will not be left holding this bag. This is such financial engineering nonsense, and if we had any sort of regulatory controls this would never be allowed to happen - especially BECAUSE of national security reasons.
This makes a lot of sense. The commercial launch business is not large enough to support all possible Falcon launches, so Starlink was created to take advantage of the low launch cost and vertical integration and is now a major profit center for SpaceX.
Starship launches are only going to make sense every 779.94 days (the approx 2 year Mars-Earth proximity). The rest of the time, the launches could similarly be used to deploy orbiting data centers for XAi/Grok etc. Brilliant move.
Do this math include the cost and weight of the radiators? Because it obviously can't work without big radiators, and I don't see them mentioned in the math?
Friendly reminder for anyone that forgot - xAI acquired Twitter, so now Space-X is the proud owner of a dying social media platform that they overpaid for.
Any claims that this is about putting compute in space is just a non-sense distraction. This was absolutely about bailing Elon out of his impulsive, drug-fueled Twitter purchase.
The only question now is: when they try to go public, will they be punished for wasting so much money or not? My guess is: not.
Reminder that SpaceX has received an estimated $38 billion in government funding over the years, and all of its returns are going to a small set of private investors.
Socialized losses, privatized profits. As is the American way.
We just had an X8.1 CME event, I just want to point out that at any moment we could have an x40 (we had a carrington event already in 1859) or higher event and all those sats at low earth orbit would be fried and start hitting each other, if SpaceX keeps launching more it becomes incredible probable that we might hit the Kessler Syndrome, and we would legit lose access to Space for a WHILE, including all of what satellites entail.
Are we ready for that as a modern society or are we going to start enacting regulation against it? I'm sorry but people wanting internet everywhere does not justify we going back to the dark ages for a decade or more.
US is dialing back even on global warming. There is no chance current government would risk space superiority for some kessler-shmessler that nobody has ever seen.
If Musk and SpaceX are serious about putting 1 million datacenter satellites into space, then they are not serious about Mars.
You cannot simultaneously build and launch 10’s of thousands of Starships to deliver 1 million tons of equipment and supplies bound to Mars while also committing to launching 10’s of thousands of Starships to orbit full of satellites.
They would need to quadruple their launch rate, and half of those launches would be Starships bound for Mars, the vast majority of which would never return.
How many Falcon9’s have ever been built? It is incredible to say you can build that many rockets and use up that much fuel on any reasonable time scale. You might as well say the Tesla Roadster version 3 will be a Single Stage to Orbit rocket car.
Pre-IPO price padding. xAI is going nowhere but at least for now it has some value. Move it under SpaceX, bump up SpaceX’s valuation and therefore it’s opening IPO price. Then kill xAI and write it off.
I really would like to see a cost and cooling breakdown. I just can't see how you can do radiative cooling on the scales required, not to mention hardening.
I thought this was a troll by Elon, now I'm leaning towards not. I don't see how whatever you build being dramatically faster and cheaper to do on land, even 100% grid independent with solar and battery. Even if the launch cost was just fuel, everything else that goes into putting data centers in space dwarfs the cost of 4x solar plus battery.
I cannot really tell satire apart from genuine opinions anymore.
(But I do hope it was satire, if not, cooling satelites was/is a big issue and they only have very modest heat creation. A data center would be in a quite different ballpark)
Now he just needs to work in crypto satellites down to users via all the new phones supporting satellite link to SpaceX. I kinda expected that one first. Distributed payment network outside of government control/oversight seems like something he would be in to.
So they use a valid and valuable company to hide a giant dumpster fire company. To add to that, their best argument is "AI in space", which has some real "solar roadways" energy to it. I honestly don't know how any SpaceX shareholder could approve this.
Tesla has probably the most valuable shareholders on Earth. Over years of empty promises and meme status, the stock has pretty much purged all the level heads. So it's mostly deluded Elon sycophants giving placing their tithe on the alter of his sci-fi fantasy smoke and mirrors game.
In reality he will be dumping the debt of twitter and xAI (and maybe spacex?) on Tesla shareholders, and buoying that with the added layer of hyper that spaceX brings.
Pretty terrible for SpaceX. Of course they paid a crazy inflated price for xAI in an attempt to cash in on the IPO. This just devalues SpaceX and exposes the investors to all the AI bubble risk.
> SpaceX has acquired xAI to form the most ambitious, vertically-integrated innovation engine on (and off) Earth, with AI, rockets, space-based internet, direct-to-mobile device communications and the world’s foremost real-time information and free speech platform. This marks not just the next chapter, but the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!
Reminder that space only allows for radiative cooling (since there is no air to absorb heat) so data centers in space are going to have massive cooling panels.
The way I read it, X AI is not really profitable and Elon's creditors/coinvestors asked for something tangible for their money, a.k.a shares in SpaceX, his only business that still has some solid foundation. The rest is emois.
Major grift vibes, inventing half-baked reasoning to justify massive valuations.
If money wasnt so deeply entwined with politics at this point, this is the sort of news that would launch fraud investigations.
Hyperloop was never a company project, neuralink was a separate company, tesla is rolling out driverless robotaxis and fsd is amazing, data centers seem, robots inevitably are going to do the majority of work - there’s no real doubt about it is there?
> robots inevitably are going to do the majority of work - there’s no real doubt about it is there?
There's a lot of doubt that the AI and compute to enable that would happen on commercially relevant timescales.
Consider: "do the majority of work" is a strict superset of "get into car and drive it". The power envelope available for an android is much smaller than a car, and the recently observed rate of improvements for compute hardware efficiency says this will take 16-18 years to bridge that gap; that plus algorithmic efficiency improvements still requires a decade between "car that can drive itself" and "android that can drive a car". (For any given standard of driving).
And that's a decade gap even if it only had to do drive a car and no other labour.
You can't get around this (for an economy-wide significant number of androids) by moving the compute to a box plugged into the mains, for the same reason everyone's current getting upset about the effect of data centres on their electricity bills.
And note that I'm talking about a gap between them, not a time from today. Tesla's car-driving AI still has safety drivers/the owner in the driving seat, it is not a level 4 system. For all that there are anecdotes about certain routes and locations where it works well, there's a lot of others where it fails.
That said: Remote control units without much AI are still economically useful, e.g. a factory in Texas is staffed entirely by robots operated over a Starlink connection by a much cheaper team in Nairobi.
Elon Musk is a genius, but he’s a financing genius. Look at the long history he has of false promises supporting financing deals between his companies and you’ll see this for what it is, a cash injection and a lie to justify it. He did the same thing with a fake solar roof demo when Tesla bought the almost bankrupt Solar City. He also shifted resources from Tesla and SpaceX to support X in the early days. Even founding xAI outside of Tesla, when so much of its valuation was built on its AI capabilities, was questionable.
I thought this wasn't viable due to cooling requirements - how do you cool massive amounts of compute when the only option is to radiate it into space - nothing to convect it with?
Also, the incredible amount of grift here with the left hand paying the right is scarcely believable. Same story as Tesla buying Solarcity. Board of directors should be ashamed IMO.
Yes. It is very cold up there but there is also no matter, or very little matter. So head conduction and convection don't work, it's all radiation. When we are learning to solve heat transfer problems in engineering school we are generally taught to neglect radiation, because it's effect on cooling the system is typically second or third order when compared to the to "big C's"
Not my experience at a hyperscaler, at least a while back. It definitely made financial sense to swap a small part to get a ~50-100k$ server's capacity back online.
When does the market realize this is all just a shell game and the emperor really has no clothes?
We saw this on a much smalelr scale a decade ago when one of Elon's companies (Tesla) acquired a second one of Elon's companies (SolarCity) because it was broke and owed a ton of money to a third one of Elon's companies (SpaceX).
Elon was forced to go through with his impulsive Twitter acquisition by a Delaware court, an acquisition that was not only secured by a bunch of Tesla stock but also a bunch of Qatari and Saudi royal money. He then mismanaged Twitter so badly Fidelity wrote down its value by at least 80% [1].
So what did Elon do? Raised even more questionable foreign money into xAI, diverted GPUs intended for another of his companies (Tesla) into Twitter and then "merged" Twitter into xAI, effectively using other people's money to bail him out from an inevitable margin call on his Tesla stock.
Interestingly, Twitter was reportedly valued at $33 billion in this deal [2], significantly more than the less than $10 billion Fidelity valued Twitter at. Weird, huh? With a competent government, this would be securities fraud that would have you spend the rest of your life in jail. And even with all that, $11 billion was lost on the deal.
So here we are and it's time for the shell game to be played again. Now it's SpaceX's turn to bail out the xAI investors.
And what is the argument for all this? AI data centers in space. Words cannot describe how little sense this makes. Launch costs (even if the Starship launch costs get to their rosy projections), cooling in space, cosmic rays (and the resulting errors) and maintenance. Servers constantly need parts replaced. You can just deorbit the satellite instead but that seems like an expensive way of dealing with a bad SSD or RAM chip.
> This marks not just the next chapter, but the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!
Making a "sentient sun" is the most bald-faced asinine, drug-induced nonsense that should be the complete destruction of all credibility to anyone who said it or typed it or works for anything here.
Financial engineering. Twitter under Elon became a dumpster fire of porn and hate and big banks were holding 13B in bonds that wouldn’t be worth the paper they were printed on for the company alone so he just links it with his only company that actually is doing something worthwhile…
Not sure how X which “merged” wit X (formerly Twitter) and SpaceX really matter or synergize but here we are. It’s all about the money being protected. And this Ketamine using wierdo is gonna be the worlds first trillionaire. Yay all of us.
> By directly harnessing near-constant solar power with little operating or maintenance costs, these satellites will transform our ability to scale compute. It’s always sunny in space! Launching a constellation of a million satellites that operate as orbital data centers is a first step towards becoming a Kardashev II-level civilization, one that can harness the Sun’s full power, while supporting AI-driven applications for billions of people today and ensuring humanity’s multi-planetary future.
Apparently optimus robots don't work and he needs to start his final grift, space datacenters, while his datacenters on earth are powered by gas turbines.
Most likely he's just trying to bury his epstein involvement where was exposed lying by his own daughter.
Slightly less high Musk aide: "But what is the synergy, where's the moat and how could that be done in practice and most importantly is there any limiting factor on Earth before we have to bring AI into.."
Musk : "SPACE!!!!!!"
It is incredible to think that the extremes of the stock market are actually pretty similar, pink sheets/cryptos and these mega companies are actually the same. News fueled pumps and dumps to win the cycle of hype of the week
wow this is one of the most incredibly fraudulent things that ever happened in American capitalism and I'm not ignoring Enron or the mortgage meltdown. I'm speechless the US has given up on any semblance of law and order in matter of financial markets and this stuff can happen without people going to jail.
There's an epic conflict of interest here with Musk owning most of both companies. And they're in entirely separate fields, there is no plausible synergy here to be gained.
Yeah but who can be hurt by this, these are both private companies? So whose interest is his "conflicting" with? I'm sure the shareholders will raise it with him and/or bring a lawsuit if they aren't happy (they probably are happy).
How can you have a conflict of interest if they're entirely separate fields? They have different interests, so where's the conflict?
You don't need synergies to justify a merger. They're often used as justification as in paying well above market price. But it has nothing to do with actual justification. You can just have a holding company of businesses
Yeah, I'm not buying that. I don't see how that could be any cheaper than regular datacenters. It might just be technically feasible, but launching stuff into space will always be more expensive than not launching stuff into space. And all those pesky technical issues like cooling might be solvable, but I doubt they're that cheap to solve.
No he is not. It makes no sense from a physics standpoint or an economic standpoint. And even if they were, it wouldn’t require whatever this acquisition is.
Musk has a history of having one of his more successful companies buying one of his less successful companies. xAI bought X, and Tesla bought SolarCity
Musk is notorious for shuffling assets across his companies to make some financials look better. For example, shuffling Twitter servers (and then all of Twitter) under xAI.
my partner got shingles a couple years ago, it was a very painful experience!
(to be crystal clear, I am making a joke equating the failed SolarCity/Tesla solar shingles to the (generally considered very painful) Herpes Zoster manifestation also called "shingles")
What's funny? Do you think the investors are against this? The investor's aren't idiots. I imagine the typical investor in Elon Musk's companies would approve of this sort of thing. So what's the problem? Besides, its a private company with Musk as majority shareholder in both. That's the beauty of private companies, you can just do things.
I wish more companies were private and ambitious. I'm tired of companies like Apple making marginal spec bumps to their phones and milking the same products for decades
More than 70% of voting shares supported the package, very close to the level of support in the original 2018 vote. This excludes Musks share.
And consider that this is retroactive, meaning it's backpay. They're literally voting to give the guy $50b for work performed. He has a lot of confidence from his investors. And if there were issues, there would be lawsuits. Ironically the only lawsuits that get brought up, like the one about the pay package, are basically trolls, from a guy that had 9 shares.
Besides the parent is the one making a claim that something not above board is going on so burden of proof is on him.
Finally, it's a private company where Musk is the majority shareholder. He's moving money from one pocket into another, and any moves will be reflected in his attempt to raise money with the IPO coming this year.
The investors want to cash out, Musk needs lots of money to plow into his latest toy that so far only excels in ridiculing him and sexual harassment/CSAM, so they make a deal to take in xAI and go public. Win win.
It's widely reported that Musk is a majority shareholder of xAI and the controlling shareholder of SpaceX (close to 80% of voting shares). Not surprising that he would be looking to consolidate ownership under one entity especially if he perceives significant synergies (i.e., data centres in space).
Shocked to see SpaceX buy the datacenter in space meme. Where does the power come from? Where does the heat go? Why add (high) launch costs to your buildout capex? Why add radiation as another risk factor to your already-unreliable GPUs? Am I missing something fundamental here...?
Not to mention the huge issues of cosmic rays. Sure, if the lifespan of the satellite is expected to be low, then maybe tolerable. But even then, how would this be financially viable?
I think it's far more likely that he wants to combine his businesses to roll his really expensive, debt-ridden companies into one entity with the company that actually reliably makes money.
Aside from Elon Musk, there are a few other people with a lot of capital aiming to do the same thing. That means, either they are all wrong (possible) or this problem has been solved somehow and the solution itself is not public.
Google and Amazon are doing the same thing. Maybe it is a moonshot (pun intended), but Musk is hardly alone in the push.
Indeed. But it's also a hilariously Musky idea! Some moderate technical competence paired with sociopathy and an ego orders of magnitude too big, and voila, you get Cybertrucks, Hyperloops, Neuralinks, Teslabots, datacenters in space, and all the other garbage the man spews.
I cannot wait for him to one day be hit in the face by reality.
I have never understood how Data centers in space ever make economic sense, the payload, latency and many other issues make it difficult at least for the immediate needs
This means that Grok, Elon's politically labotomized involuntary pornography generator, and X (formerly Twitter), Elon's Nazi-adjacent propaganda machine, are now completely intertwined with SpaceX, a too-big-to-fail government contractor that currently serves as America's only reliable option for manned orbital spaceflight.
Anyone who doesn't see how broken this situation is isn't paying attention. This is how people like Elon, who want to seize as much power from the government as they can, ensure that the means for seizing that power are untouchable.
Anyone who has ever used Grok or X lately knows that both of these products are heavily manipulated to align with the political, social, and economic views of Elon Musk, who is increasingly boosting "white power" language and full-throatedly backing America's most nationalistic and authoritarian president to date.
This is just another consolidation of power, and it's deeply worrisome. Any integrity one may have hoped remained at SpaceX just vanished when they aligned their mission with that of these deeply problematic digital services.
And this is not even scratching the surface of what looks like a deliberate attempt to create Kessler syndrome by launching millions of cheap short-term satellites into orbit, or the rationality of putting datacenters into orbit in the first place...
I hate to be that guy that posts an AI-generated response, but I thought Grok had an interesting take on this problem:
Energy Generation Practicality
Building satellite datacenters powered by solar energy in space is conceptually feasible, as solar insolation in orbit is about 1,366 W/m²—roughly 35% higher than on Earth's surface due to no atmospheric filtering or day-night cycles in geostationary or sun-synchronous orbits. Advanced solar panels in space can achieve efficiencies of 25-35%, yielding 340-480 W/m² of electrical power. For context, a typical terrestrial datacenter rack consumes 8-15 kW, but AI/high-performance computing (HPC) racks can reach 30-80 kW. Scaling to a hypothetical 1 MW datacenter (enough for ~12-30 AI racks) would require approximately 2,000-3,000 m² of solar panels, assuming 25% efficiency and some losses.
However, practicality hinges on mass and deployment. Thin-film solar arrays (e.g., from projects like Caltech's Space Solar Power Project) can weigh as little as 0.5-1 kg/m², making the array for 1 MW ~1-3 tons. This is manageable for a constellation, but integrating with compute hardware adds complexity. Concepts like Google's Project Suncatcher and Thales Alenia Space's ASCEND study propose modular satellites with integrated solar power, potentially achieving 10x lower carbon emissions than Earth-based datacenters. Challenges include orbital shading in low Earth orbit (LEO) and panel degradation from radiation/micro-meteoroids, but overall, generating sufficient energy is practical with current tech if scaled modestly (e.g., 100-500 kW per satellite).
Heat Radiation Practicality
Heat dissipation in space relies solely on thermal radiation, as there's no atmosphere for convection or conduction to a medium. This is highly practical because space acts as an infinite cold sink (~3 K background temperature). Waste heat from datacenter compute (typically 80-90% of input power) can be radiated via large, lightweight panels with high emissivity coatings (ε ≈ 0.9).
Using the Stefan-Boltzmann law for blackbody radiation (P = εσA(T⁴ - T_sink⁴), where σ = 5.67 × 10⁻⁸ W/m²K⁴), a radiator at 50°C (323 K) emits ~550-620 W/m² (one-sided). For a 1 MW heat load, you'd need ~1,600-1,800 m² of radiator area—comparable to solar panel sizing and feasible with deployable structures like those on the ISS (which radiate up to 350 W/m²). Designs from Laird Technologies and NEC emphasize distributed radiation at the PCB level, using materials like aluminum or composites for efficient emission while minimizing solar absorption.
The main limitation is surface area: satellites must balance radiator size with stability and drag in LEO. Overheating risks exist during eclipses or high-load bursts, but passive systems (e.g., heat pipes) and active louvers make this solvable. Studies like ASCEND confirm space datacenters could reject heat more efficiently than Earth-based ones, avoiding water-intensive cooling.
Shielding from Cosmic Rays in LEO
In LEO (400-2,000 km), Earth's magnetic field deflects much of the high-energy radiation, resulting in dose rates of 100-10,000 rad(Si)/year—far lower than in higher orbits. This makes shielding practical and often minimal for short missions (1-5 years). Primary threats are total ionizing dose (TID) causing degradation and single-event effects (SEE) like bit flips or latch-ups.
Standard aluminum shielding (3-5 mm thick, ~1-2 g/cm² areal density) reduces TID to <10 krad(Si) over 3 years, sufficient for 65-130 nm chips. Polymers like polyethylene or borated polyethylene outperform aluminum by 20-70% in proton shielding, reducing displacement damage. Many LEO cubesats use commercial off-the-shelf (COTS) electronics with redundancy (e.g., multiple computers) and error-correcting code, avoiding heavy rad-hard parts that add mass/cost.
For datacenters, radiation-tolerant designs (e.g., from Melagen Labs' low-density composites) can extend COTS lifespan 3-5x. No exotic shielding is required—Earth's magnetosphere provides natural protection, and simulations (e.g., MCNP/OMERE) confirm viability. Drawbacks include occasional upsets during solar flares, but overall, LEO shielding is straightforward and cost-effective.
Launch Costs for Practical Mass and Competitive Compute
Launch costs are the biggest barrier. Current prices to LEO are ~$2,700/kg (Falcon 9) to ~$1,500/kg (Falcon Heavy). For a "useful" compute level—say, 1 petaFLOP/s (comparable to a small AI cluster)—hardware might mass 1-5 tons (based on dense racks like NVIDIA DGX at ~100-200 kg/rack, plus power/thermal systems). Including solar panels and radiators, a single satellite could total 5-10 tons.
At $1,500/kg, launching one unit costs $7.5-15 million—potentially competitive with terrestrial datacenter energy costs ($0.05-0.15/kWh over lifetime), but scaling to a constellation (10-100 satellites for global coverage) pushes totals to $75-1,500 million. Google's analysis requires costs <$200/kg by the mid-2030s for parity; Starship aims for $10-100/kg, dropping per-satellite costs to <$1-2 million.
Competitiveness: Terrestrial datacenters cost ~$10-20 million/MW to build/operate annually, but space avoids land/water constraints. At current costs, space is impractical for mass deployment; with Starship, it could undercut Earth prices by 10-50x in energy efficiency. Thales estimates needing 10x less emissive launchers for net-zero viability.
Overall, the concept is technically practical today for prototypes (e.g., via Starcloud or SpaceX proposals), but economically viable only with launch costs dropping 10-20x. Energy and heat are strengths; shielding is low-risk in LEO. Full constellations could emerge by 2030 if reusability scales.
People pooh-poohing space datacenters will obviously think this is a bad move. But Elon clearly believes space datacenters will work. Given that, and the fact that SpaceX will IPO this year, this acquisition was inevitable.
SpaceX and xAI would not be able to freely collaborate on space datacenters after the IPO because it would be self-dealing. SpaceX likes to be vertically integrated, so they wouldn't want to just be a contractor for OpenAI's or Anthropic's infrastructure. Merging before the IPO is the only way that SpaceX could remain vertically integrated as they build space datacenters.
From the Big Short (movie)
Jared Vennett (narration): "In the years that followed, hundreds of bankers and rating agency's executives went to jail. The SEC was completely overhauled, and Congress had no choice but to break up the big banks and regulate the mortgage and derivatives industries."
"Just kidding. Banks took the money the American people gave them, and they used it to pay themselves huge bonuses, and lobby the Congress to kill big reform. And then they blamed immigrants and poor people, and this time even teachers."
I feel like without adding some commentary with these quotes this comment lacks enough info to see how it relates to the linked article.
Because this move is entirely financial engineering to hide losses just like the roll up of X in to xAI.
None of this has anything to do with business or innovation. Do you not immediately see that? Most of my friends reaction to this news was that this is so obvious it's almost funny (or actually it is funny, since most were laughing as they read the headline).
I'm curious how you could not understand the relevance of the quote unless you were aggressively trying to not understanding it.
> I'm curious how you could not understand the relevance of the quote unless you were aggressively trying to not understanding it.
Monet probably wondered how other people couldn't see purple in a haystack.
Parent poster may have been thinking of other readers. I see it as you do, but it's a fair question.
yeah, I am not a huge fan of Musk, but this move is just going to bring down arguably the only decent thing he's produced.
Leave SpaceX alone you child. Gwynne has it in excellent hands.. find some other way to pay for your juvenile brainfarts.
Legitimately, did you not immediately conclude it was for financial shenanigans? What did you think? I'm not trying to be shitty, but what else could there be?
You don't see how the common thread is Elon Musk buying out his own businesses, largely on the basis of overinflated stocks and corporate welfare?
It's baffling that the market has stayed so irrational because of Musk. It will collapse because of him.
This merger smells like a bubble. Servers in space? They don't make enough to cover costs here on earth. Americans will be forced to bail this mess up because we need Falcon 9, Starship one, etc.
Agree, unless your IQ is over 10 and you don't live under a rock.
Then you're a fool
It's messed up that Grok underwrote all those subprime mortgages in 2008
I think the argument is that it's messed up that a large debt swap from xAI kept Musk's margin on Twitter from being called by his investors, and now that debt is being absorbed by SpaceX.
> Musk's margin on Twitter from being called by his investors,
Primary and largest investors in X are: Elon Musk, Saudi Prince Alwaleed bin Talal, Larry Ellison, Jack Dorsey.
I don't know that you need to worry about their financial well-being or that they are getting a raw deal.
I think people are more concerned about SpaceX getting the raw deal here.
And specifically that if the music is about to stop SpaceX has an implicit government backstop
Yeah, the financial well-being of those investors is not what people are worried about here
I really don't know what you're trying to say. From your comment alone the conclusion I drew was that we should spend more on the industries that makes physical instead of financial products, such as SpaceX.
The sequence of events: Elon doing a leveraged buyout of X, then xAI funding, then debt transfers to X, then the xAI–X stock deal. Now the proposed SpaceX–xAI merger appears to have shifted X’s financial burden from Musk personally toward xAI investors and, potentially, future SpaceX shareholders.
This is speculative, of course, but yeah seems likely.
So what? People who buy SpaceX shares should take into account all of its debt when deciding how much to pay for a share.
SpaceX is a private company
so are the banks?
>> SpaceX is a private company
> so are the banks?
Which relevant bank do you have in mind that is not a public company (listed on a stock exchange)?
Where's the government bailout in this transaction that would make this a relevant comparison?
Who is SpaceX’s biggest customer? And which industry are we being told by any number of governments around the world is the most importantly thing ever and must be subsidized and forced on people?
> Who is SpaceX’s biggest customer?
It is estimated that Starlink is, accounting for 70% - 80% of revenue. Sources: [1] and [2]
NASA is SpaceX's biggest external customer for rocket launch services.
Although NASA is SpaceX’s largest external customer for traditional launch services, the company earns far more revenue from Starlink customers (millions of subscribers). So overall Starlink itself is SpaceX’s biggest revenue generator and de facto largest customer segment.
[1] https://pestel-analysis.com/blogs/target-market/spacex
[2] https://londoneconomics.co.uk/blog/publication/crouching-riv...
SpaceX saves its biggest customer money by being the superior, cheaper launch option. The alternative was ULA, which was an extraordinarily expensive monster.
Please highlight the problems you have with how it pertains to this context, how the biggest customer is harmed.
What do you care if its private owners are willing to absorb the mess that is xAI?
There has been rumblings of speculation that when the AI bubble pops that the government will bail the companies out.
> it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power
We currently make around 1 TW of photovoltaic cells per year, globally. The proposal here is to launch that much to space every 9 hours, complete with attached computers, continuously, from the moon.
edit: Also, this would capture a very trivial percentage of the Sun's power. A few trillionths per year.
Context missing. This is in reference to a vision the (distant?) future where the satellites are manufactured in factories on the Moon and sent into space with mass drivers.
Full paragraph quote comes from:
> While launching AI satellites from Earth is the immediate focus, Starship’s capabilities will also enable operations on other worlds. Thanks to advancements like in-space propellant transfer, Starship will be capable of landing massive amounts of cargo on the Moon. Once there, it will be possible to establish a permanent presence for scientific and manufacturing pursuits. Factories on the Moon can take advantage of lunar resources to manufacture satellites and deploy them further into space. By using an electromagnetic mass driver and lunar manufacturing, it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power. >
Help me understand something. We make 1 TW of cells per year but we're struggling with bringing 1 GW consuming data centers online?
> We currently make around 1 TW of photovoltaic cells per year, globally.
China made 1.8 TW of solar cells in 2025.
The raw materials required to make these are incredibly abundant, we make as much as we need.
you realize the factor of 2 you introduce doesn't meaningfully change the order of magnitude that the previous poster is implying right?
You missed the point.
We can make ten or hundred times the number of solar cells we make right now, we just don't have a reason to. The technology is fairly ancient unless you want to compete on efficiency, and the raw materials abundant.
See Dyson Sphere: https://en.wikipedia.org/wiki/Dyson_sphere
Dyson's paper was literally written in jest.
Yeah, that's the point ... it's stupid to believe humanity is capable of deploying that much infrastructure. We cannot do even 0.01% of it.
What do you think the limiting factor is? I don't see why we can't scale manufacturing of satellites up as far as we want. If we mine out a substantial fraction of the mass of the earth, we can go harvest asteroids or something.
>> Dyson Sphere
> What do you think the limiting factor is?
You need to be able to harness enough raw material and energy to build something that can surround the sun. That does not exist in the solar system and we do not yet have the means to travel further out to collect, move, and construct such an incredibly huge structure. It seems like a fantasy.
There are only so many people who can make satellites; there are only so many things to make satellites out of; and there are only so many orbits to put them in. There are only so many reasons why a person might want a satellite. There are only so many ways of placing satellites in orbit and each requires some amount of energy, and we have access to a finite amount of energy over time.
Finally, if we limited ourselves to earth-based raw materials, we would eventually reach a point where the remaining mass of the earth would have less gravitational effect on the satellite fleet than the fleet itself, which would have deleterious effects on the satellite fleet.
Seven reasons are intuitive; I’m sure there are many others.
After a few decades, you need to start replacing all the solar panels.
And the robot army being used to do the construction and resource extraction will likely have a much shorter lifespan. So needs to be self-replicating/repairing/recycling.
The physical amount of material in the solar system is a pretty big limiting factor.
Yeah, but besides not having the amount of having the physical amount of material available in the solar system, or the availability of any technology to transfer power generated to a destination where it can serve a meaningful purpose in the foreseeable future, or having the political climate or capital necessary for even initiating such an effort, or not being able to do so without severely kneecapping the habitability of our planet, there are aren't really any meaningful barriers that I can see.
But the factory ~~can~~must grow.
In fairness, solar cells can be about 5x more efficient in space (irradiance, uptime).
The quoted "1 TW of photovoltaic cells per year, globally" is the peak output, not the average output. They're only about 20% higher peak output in space… well, if you can keep them cool at least.
But there are no clouds in space and with the right orbit they are always facing the sun
The 1TW is the rated peak power output. It's essentially the same in space. The thing that changes is the average fraction of this sustained over time (due to day/night/seasons/atmosphere, or the lack of all of the above).
It's still the same 1TW theoretical peak in space, it's just that you can actually use close to that full capacity all the time, whereas on earth you'd need to over-provision substantially and add storage, so 1TW of panels can only drive perhaps a few hundred GW of average load.
> the whole capacity
Wouldn’t something like half of the panels be in shadow at any time?
It is more than 5x less expensive to get surface area on earth’s surface.
The dominant factor is "balance of system" aka soft costs, which are well over 50%.[0]
Orbit gets you the advantage of 1/5th the PV and no large daily smoothing battery, but also no on-site installation cost, no grid interconnect fees, no custom engineering drawings, no environmental permitting fees, no grid of concrete footers, no heavy steel frames to resist wind and snow loads. The "on-site installation" is just the panels unfolding, and during launch they're compact so the support structure can be relatively lightweight.
When you cost building the datacenter alone, it's cheaper on earth. When you cost building the solar + batteries + datacenter, it (can be) cheaper in space, if you build it right and have cheap orbital launch.
[0] https://en.wikipedia.org/wiki/Balance_of_system
[delayed]
No maintenance either
Right now it is.
However, the amount of available land is fixed and the demand for its use is growing. Solar isn't the only buyer in this real estate market.
We have so much excess land with no real use for it that our government actually pays farmers to grow corn on it to burn in cars.
Availability of land for solar production isn't remotely a real problem in the near term.
The Technology Connections Youtube channel recently did a great video arguing pretty convincingly that the land used to grow corn for cars would be vastly more efficiently used from an energy perspective if we covered it with solar panels.
This.
I feel like everyone just lost their mind.
You just have to remember, most of these people live in high density regions and have little comprehension about how much surface area humanity truly occupies... And that isn't even accounting for offshore constructs.
Realizing the impracticality of it (and that such approaches often collapse under the infeasibility of it) ... wouldn't it be better to... say... cover the Sahara in solar panels instead? That's gotta be cheaper than shipping them into space.
https://inhabitat.com/worlds-largest-solar-project-sahara-de...
https://www.theguardian.com/business/2009/nov/01/solar-power...
(and a retrospective from 2023 - https://www.ecomena.org/desertec/ )
Solar can always just go on the roof...
the demand is pretty much fake and AI isn't actually making money, just gobbling investors money
Fortunately there are no downsides to launching solar cells into space that would offset those gains.
Does that include all the required radiators to vent heat?
Solar cells have exactly the same power rating on earth as in space surely? What would change is their capacity factor and so energy generation.
Satellites can adjust attitude so that the panels are always normal to the incident rays for maximum energy capture. And no weather/dust.
You also don't usually use the same exact kind of panels as terrestrial solar farms. Since you are going to space, you spend the extra money to get the highest possible efficiency in terms of W/kg. Terrestrial usually optimizes for W/$ nameplate capacity LCOE, which also includes installation and other costs.
For one or a few-off expensive satellites that are intended to last 10-20 years, then yes. But in this case the satellites will be more disposable and the game plan is to launch tons of them at the lowest cost per satellite and let the sheer numbers take care of reliability concerns.
It is similar to the biological tradeoff of having a few offspring and investing heavily in their safety and growth vs having thousands off offspring and investing nothing in their safety and growth.
Solar modules you can buy for your house usually have quoted power ratings at "max STC" or Standard Testing Conditions, which are based on insolation on Earth's surface.
https://wiki.pvmet.org/index.php?title=Standard_Test_Conditi...
So, a "400W panel" is rated to produce 400W at standard testing conditions.
I'm not sure how relevant that is to the numbers being thrown around in this thread, but thought I'd provide context.
The atmosphere is in the way, and they get pretty dirty on earth. Also it doesn't rain or get cloudy in space
Sure but like, just use even more solar panels? You can probably buy a lot of them for the cost of a satellite.
The cost of putting them up there is a lot more than the cost of the cells
“This scale” is not realistic in terms of demand or even capability. We may as well talk about mining Sagittarius A* for neutrons.
And in geostationary, the planet hardly ever gets in the way. They get full sun 99.5% of the year.
even at 10% (say putting it on some northen pile of snow) it is still cheaper to put it on earth than launch it
I don't disagree
I'm all for efficiency, but I would think a hailstorm of space junk hits a lot harder than one of ice out on the farm.
Except it doesn't melt like regular hail so when further storms come up you could end being hit by the same hail more than once :\
Atmospheric derating brings insolation from about 1.367KW/m2 to about 1.0.
And then there’s that pesky night time and those annoying seasons.
It’s still not even remotely reasonable, but it’s definitely much higher in space.
> And then there’s that pesky night time and those annoying seasons.
The two options there are cluttering up the dawn dusk polar orbit more or going to high earth orbit so that you stay out of the shadow of the earth... and geostationary orbits are also in rather high demand.
Put them super super far away and focus all the energy into one very narrow death laser that we trust the tech company to be careful with.
Now do waste heat.
And how much of that power would be spent on high speed communications with Earth that aren't, you know, a megabit or two per second
That's pretty much a solved problem. We've had geostationary constellations for TV broadcast at hundreds of megabytes for decades now, and lasers for sat-to-sat comms seems to be making decent progress as well.
> it is possible to put 500 to 1000 TW/year of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the Sun’s power
Which satellites are operating from "deep space"?
Those are for video. AI Chat workflows use a fraction of the data.
I grew up on a rural farm in California with a dial-up connection that significantly hampered my ability to participate in the internet as a teenager. I got Starlink installed at my parents' house about five years ago, and it's resulted in me being able to spend considerably more time at home.
Even with their cheapest home plan, we're getting like 100 Mbps down and maybe 20 to 50 up. So it's just not true at all that you would have connections that are a megabit or two per second.
That's not what I'm suggesting. The post says "deep space". If you're going to try to harvest even a tiny percentage of the sun's energy, you're not doing that in Earth's orbit. The comparison is a webcam feed from Mars.
Pfft that would just require setting up an entire lunar mineral extraction and refining system larger than we have on earth, just minor details.
This is quite frankly the largest amount of bollocks Musk has come up with so far. The market is basically a meme market:
Same some futuristic sounding shit -> stonks go up -> repeat
All literally while trying to do really bad things everywhere which are going to end up with bodies piled to the fucking ceiling. And not delivering products. And being a proto-nazi.
This needs to break. Now. No one should be rewarding this shit.
> And not delivering products
2024 revenue of >$100b is quite impressive for not delivering any products
You know what they mean. Full self-driving was promised what, 10 years ago? Tesla Roadster? Sub-25K car? etc etc etc
I should say delivering promised products.
Anyway they just canned the S and X lines so that's done as well...
What kind of nonsense is that. SpaceX 2024 revenue barely broke $10B, if that. Launch was probably ~$4B and Starlink probably ~$5B. I'll give them the benefit of the doubt and double those just for shits and giggles and that's still less than $20B and you're claiming >$100B? Horse shit. Nonsense.
:eye roll:
This schtick is so, so tiresome.
What is tiring is people defending it. Everyone goes down with this ship...
A former NASA engineer with a PhD in space electronics who later worked at Google for 10 years wrote an article about why datacenters in space are very technically challenging:
https://taranis.ie/datacenters-in-space-are-a-terrible-horri...
I don't have any specialized knowledge of the physics but I saw an article suggesting the real reason for the push to build them in space is to hedge against political pushback preventing construction on Earth.
I can't find the original article but here is one about datacenter pushback:
https://www.bloomberg.com/opinion/articles/2025-08-20/ai-and...
But even if political pushback on Earth is the real reason, it still seems datacenters in space are extremely technically challenging/impossible to build.
We don’t even have a habitable structure in space when the ISS falls, there is no world in which space datacenters are a thing in the next 10, I’d argue even 30 years. People really need to ground themselves in reality.
Edit: okay Tiangong - but that is not a data center.
Who is “we”? https://en.wikipedia.org/wiki/Tiangong_space_station
Good point. Still a long, long way from data centers.
I don't think any of the companies that say they are working on space data centers intend them to be habitable.
Ok then short SpaceX stock when it IPOs.
What does stock price have to do with anything?
That someone could put a data center in space for the price of 100 years of eliminating world hunger doesn’t mean shit.
"Technically challenging", a nice way to say "impossible"
Just like rockets landing themselves
SpaceX is too big to fail. It's important for national security.
I wonder if Elon wants to tangle all his businesses into SpaceX so they are all kept afloat by SpaceX's importance.
Elon can't legally financially entangle Tesla to SpaceX due to Tesla being a public company, so his hands are tied.
Tesla is clearly benefiting from protectionism and its sales would collapse if BYD were allowed to openly sell in the US. Most people just want affordable, maintainable and reliable cars.
I’m old enough to remember when this was said about Solar City
Got that double digit age locked down I see, congrats!
> Elon can't legally financially entangle Tesla to SpaceX due to Tesla being a public company, so his hands are tied.
He absolutely could do it, just like he did when Tesla bought SolarCity. It just isn’t as easy when one of the companies is public than when both are private.
Tesla to invest $2B in Elon Musk’s xAI https://techcrunch.com/2026/01/28/tesla-invested-2b-in-elon-...
> Elon can't legally financially entangle Tesla to SpaceX
Bill Ackman has proposed taking SpaceX public by merging it with his Pershing Square SPARC Holdings, distributing 0.5 Special Purpose Acquisition Rights (SPARs) to Tesla shareholders for each share held. Each SPAR would be exercisable for two shares of SpaceX, aimed at enabling a 100% common stock capitalization without traditional underwriting fees or dilutive warrants.
With SpaceX IPO set to be one of the biggest of all time, this could have a pretty gnarly financial engineering impact on both companies -- especially if the short interest (direct or through derivatives) remains large.
Why would SpaceX go public? They already have a robust enough private market to give liquidity to all of their employees and shareholders who want it. They can get more private investment.
Going public would add a lot of hassle for little to no gain (and probably a negative of having to reveal their finances).
It has been widely reported for weeks that SpaceX is planning to go public in a few months. The reason is they have big plans to run a vast network of AI servers in orbit and will need to raise a massive amount of funding. xAI merger fits with that plan. I'd assume SpaceX still plans to go public.
Was ignored on HN but here's an article explaining:
https://arstechnica.com/space/2025/12/after-years-of-resisti...
it wasn't ignored on HN, there were many articles correctly noting that building data centers in space is a stupid stupid idea because cooling things there is infeasible
lol WHAT?
AI datacenters are bottlenecked by power, bandwidth, cooling, and maintenance. Ok sure maybe the Sun provides ample power, but if you are in LEO, you still have to deal with Earth's shadow, which means batteries, which means weight. Bandwidth you have via starlink, fine. But cooling in space is not trivial. And maintenance is out, unless they are also planning some kooky docking astromech satellite repair robot ecosystem.
Maybe the Olney's lesions are starting to take their toll.
Weirdest freaking timeline.
The shadow thing can be solved by using a sun-synchronous orbit. See for example the TRACE solar observation satellite, which used a dawn/dusk orbit to maintain a constant view of the sun.
Cooling, on the other hand? No way in hell.
The cooling is the bit where I'm lost on, but it will be interesting to see what they pull off. It feels like everyone forgets Elon hires very smart people to work on these problems, it's not all figured out by Elon Musk solely.
He's broken pretty much all the other financial rules.... for example, the amount of blatant self-dealing he gets away with is staggering.
As long as the consequences of his actions continue to increase the paper value for investors, regulations don't really have teeth because there aren't damages. So the snowball gets bigger and the process repeats.
It's "ironic?" considering Tesla launching in China is what created the necessary supply chain to turn BYD into the powerhouse it is today. Tesla's greed will become their own demise.
Tesla cars made in Shanghai are sold in Europe and other places. That is helping them be competitive and they haven't had much price pressure until recently. Just because the Chinese have their own internal competition and deflation which drove their prices down aggressively doesn't mean it was a bad idea to build there. Also the idea the Chinese couldn't figure it out without an American company coming there first to show them is pretty silly.
Tesla Shanghai opened in 2019
BYD made their first hybrid in 2008 and they were a battery company since the 90s
BYD are just affordable and maybe reliable, regarding maintenance their spares are hard to come by and are almost as hard to work with as Tesla and other brands.
I've done plenty of work on my own Tesla. It's not hard to work on at all. Parts are not even very difficult. There are plenty of 3rd party shops (such as one I went to when I needed to replace my windshield.) I really wonder why people continue to think this. It's not 2016 any more.
Tesla body work is extremely expensive. Aluminum, extensive welding instead of fasteners, substantially reduced modularity due to castings, specialized tooling just off the top of my mind.
Are you a car mechanic living in China?
Presumably "hard to come by" would be somewhat irrelevant in any jurisdiction other than the US?
[Nearly] all is possible when you have a board of simps/cultists
Starship has a large number of critical milestones coming: Can it land and quickly reuse the upper stage? If not, it can't make refueling flights without building a dozen or two starships. Can it carry the full specified payload? If not, it can't even try to refuel in orbit. If it can't refuel in orbit, it can't go beyond earth orbit. Etc.
Everything has to go right or it will be irrelevant before it works.
> Everything has to go right or it will be irrelevant before it works.
Starship is not all of SpaceX. Saying, maybe because one hates Musk, that SpaceX is going to become irrelevant is wishful thinking.
In 2025 SpaceX launched more rockets into space than the entire world ever sent in a year up to 2022, something crazy like that.
Then out of, what, 14 000 active satellites in space more than half have been launched by SpaceX.
SpaceX is, so far, the biggest space success story of the history of the human race (and GP is right in saying that SpaceX is now a national security matter for the US).
> In 2025 SpaceX launched more rockets into space than the entire world ever sent in a year up to 2022, something crazy like that.
Not just that, the cost of each rocket launch is drastically cheaper than all of its competitors costs.
Model S was the most successful EV. If you think cybercab is the vehicle of the future, look at the timeline of the only robo taxi in commerce in the US.
Everything has to go right with that, or cybercab will be irrelevant before it works. Same deal. Same bullshitter.
Model S was successful until Model 3/Y blew it out of the water. Waymo’s timeline is not relevant because they lose money on every car and every deployment. Tesla’s the only financially successful developer of self driving. They can scale it up much faster.In fact, instead of making $5k per car produced, cybercab will net them $50k per car per year.
The world doesn't consist of just Waymo and Tesla, and even if it did there's no guarantee either succeeds.
> cybercab will net them $50k per car per year.
Assuming no mass boycotts, nor targeted vandalism. We've already seen both in the last 12 months.
It’ll be fine. Especially when people compare the price of ownership/uber to robotaxis.
When people actually compare prices, they note that Chinese cars also have autopilot and cost less than half of a Tesla, new.
What's keeping Chinese brands out of the USA, isn't keeping them out of Europe or much of anywhere else.
97% of their sales are model 3 / Y
I think he will spin Tesla off since electrification and autonomy are no longer cool (he can’t build good quality cars or reliable FSD)
Haven't you heard? Tesla is pivoting to building humanoid robots instead. They haven't sold a single one, but it toootally warrants retooling their car factories, pinky promise!
FSD is incredibly reliable. Build quality of US built cars is middle of the pack, Europe/China built Teslas are top of the pack.
Incredible shilling, bravo
Oh c'mon now. Damn model 3 and model S I have driven were considerably lower quality interiors than an ass end Citroen or Fiat. The Model S, a 2023 model the doors didn't even fit properly. And that was all Europe.
As for FSD, nope. Unless you redefine the word reliable.
Edit: I owned a 2018 Model S as well. Literally the worst fucking car I have ever owned or driven.
I disagree. Model 3 has soft touch everywhere. Freaking bmw 3 series has plastic on most frequently touched bits.
Since you are in europe you have no idea how good fsd is.
BMW actually has a reasonable control surface though, not a grand user interface experiment by some crack heads.
As I'm in Europe I just get trains.
> BMW actually has a reasonable control surface though, not a grand user interface experiment by some crack heads.
Really? It's one thing to hate Elon Musk, but you're talking about a lot of brilliant engineers who worked on these cars, everything from the components to the software. It's uneeded low blow just because you don't like Elon Musk.
The bmw interface is the actual fucking joke. Everything you need on Teslas is accessible from the steering wheel in addition to the touchscreen.
Apart from the speedometer which is outside your safe FOV in the Tesla.
And everything in the BMW you should be dealing with when driving is on or around the steering wheel.
That’s not how human eyes work. https://en.wikipedia.org/wiki/Vision_span
> That’s not how human eyes work. https://en.wikipedia.org/wiki/Vision_span
The article you linked agrees with me. Greatest resolution in the macula which is a span of approximately 6 degrees from the centre.
Sigh...
Yeah. You’re never looking at the road and reading your speed at the same time.
SpaceX is slated to go public some time this year - June IIRC
The biggest selling point /was/ that Musk was being managed there, he wasn't tinkering with SpaceX like Twitter or Tesla, and his foolhardy direction was kept out of the company.
BUT, like Tesla, Musk cannot help himself and is making SpaceX look like a very bad investment - tying his other interests with SpaceX, allegedly using SpaceX money as a "war chest" in his battles.
There is also a danger that investors will see xAI as politically dangerous, which will really hurt SpaceX IPO
They want to go public, but have to sell the hell out of it in the meantime.
I'll bet SpaceX financials aren't as great as some people think. Remember, Elon was the guy who tried to take Tesla private, and talked a lot of smack about how silly it is to be a public company. All of a sudden he wants SpaceX to go public?
Musk has a pattern here - he used Tesla the same way, diverting resources to xAI and treating it as a funding vehicle for other ventures. Once he started doing that, Tesla's financials got murky and harder to trust. Now he's doing it with SpaceX right before the IPO. For investors, that's not 'too big to fail' protection - it's a red flag that the company finances are entangled with his personal empire instead of focused on the core business.
> The biggest selling point /was/ that Musk was being managed there, he wasn't tinkering with SpaceX like Twitter or Tesla, and his foolhardy direction was kept out of the company
The biggest selling point to who? Definitely not wall street
Let’s be honest - this is just a way to prop up Twitter/X. It makes SpaceX shareholders subsidize X, and also American taxpayers who are giving contracts to SpaceX for highly sensitive things. The government should ideally refuse to give SpaceX work unless it unwinds this.
Why? The government is paying less for SpaceX than alternatives. It th cheapest and best service.
Because Twitter/X is distorting our politics (with ann unbalanced scheme of censorship / amplification / suppression) and destroying the country by mainstreaming far right supremacist politics. Twitter/X does not deserve a single dollar of taxpayer money. If SpaceX is now part of that machine, it doesn’t deserve a single dollar either. I would rather pay more for alternatives and encourage their growth. I also look at any money given to this company as the equivalent of GOP campaign funding, so I feel it should be treated as illegal under the law.
Shouldn't the government be aiming to pay the lowest price for the best goods and services rather than using procurement as a way to promote or suppress certain political opinions?
The government is prevented from doing that by a little thing called the first amendment. "Mainstreaming far right supremacist politics" is just a hyperbolic way of saying he has politics you don't like and is exercising his freedom of the press by promoting it on the media platform he owns. Legally that is no different then the rights that every newspaper and TV station in the country has.
Musk is, indeed, allowed under the 1st to promote whatever he wants to promote. Him being a hypocrite about "free speech absolutism" is not a crime.
However, the current US administration appears to be actively violating the 1st and 5th in a bunch of ways, the 14th that one time, and making threats to wilfully violate the 2nd for people they don't like and the 22nd to get a third term.
I disagree. He would be using taxpayer money to boost his preferred speech. And it is essentially campaign funding for the GOP. It should be treated as such.
I think that line of argument would work in my country of birth, the UK, but I don't think it works in the USA.
Unsold rather our government not get in the habit of violating the multiple laws put in place to keep it from playing favorites and picking winners.
And our tax dollars.
national security is pretty felixaeble
Being too big to fail is not really a desirable outcome, it’s just better than failure.
Boeing is too important to fail as well but it’s been terrible as a shareholder
> > Boeing is too important to fail as well but it’s been terrible as a shareholder
Your opinion on Boeing being terrible as a shareholder vis-a-vis Tesla would be completely reversed if dividends and capital gains of the 2 companies were to be offered in the form of miles to be flown on Boeing planes and miles on Teslas Uber/Taxi/Autonomous taxis instead of dollars
The absolute overperformance on the stock market that Tesla has enjoyed vis-a-vis Boeing is not rooted in a concrete and tangible quality of life improvement for citizens. Not American citizens, nor global citizens for that matter.
It is my opinion that for all public companies in which it is possible to do so government should mandate payment in kind to all shareholders and board members to prevent the excessive promotional , cult and all around BS aspect of marketing to take over and allow people to profit just by riding off those, and Musk is the GOAT at that.
Why? Let it fail. Bring back NASA.
What do you mean "let it fail?" SpaceX has the most profitable launch system in the world and now operates >50% of all satellites in orbit. They aren't exactly in need of a bailout.
So proof of profitability is that they can shoot their own satellites into orbit?
When a company is operating at a scale where you are making orders of magnitude more orbital launches than NASA, operating a constellations of 10,000+ satellites, providing internet access to 10s of millions of people and 1 army, has raised $10s of billions in private markets at valuations in the $100s of billions, then the burden of proof is on you claiming the opposite.
The proof is that they are continuing to launch more mass into orbit than any other entity on the planet - while holding share liquidity events for their employees multiple times a year where they buy back shares. Proof is that they charge a lower cost to orbit than any of their competitors and has done so for years now.
Their revenue from Starlink is slated to be bigger than the entire NASA budget this year.
You want to bring back the biggest loser? NASA kept missing deadlines for 30 years
I am no fan of Musk the man. SpaceX is a strong company and Falcon is a solid vehicle. There is not a lot of competition, and NASA trying to in-source design and supply and construction of a new, reusable LEO rocket would be a complete nightmare.
I root for a competitive rocket market, but SpaceX is at the moment critical.
> Bring back NASA.
NASA is still here. Unfortunately, NATA fell victim to enshitification by government contracting. NASA even if it wants to simply cannot today design and launch a rocket. :(
NASA just splurges money. The private sector is far better when it comes to money.
... as we can tell by whatever the everloving fuck is going on with this press release.
I'm not talking specifically about SpaceX, although historically the cost of their rockets have been much lower than NASA. I'm being much more general. The public sector doesn't have the same incentives that private companies have, whether it's rockets or any other technology. It's sad, but it's the truth.
We have absolutely no way of gauging this until after SpaceX goes public.
SpaceX can use the same booster 30 times. NASAs new rocket can use it one time. We don't need to see financial statements to figure this one out.
I wouldn't be too sure. Depends on NASAs mission profiles and a lot of factors. Falcon heavy can bring 26.7t to GTO in expendable mode and only 8t in reusable mode. Reusable cost of Falcon is US$97 million vs US$150 million expendable.
How much does it cost to develop and maintain the reusability? Is it worth the trade-offs in lower tons to orbit due to more weight? Is it worth it adjusting the payload into smaller units, including developing things like refueling in LEO?
Idk, I'm not on the inside doing those calculations...
[Absolutely](https://www.reuters.com/business/finance/spacex-generated-ab...), [no way](https://payloadspace.com/estimating-spacexs-2024-revenue/), indeed.
Everything is estimated.
If you want to trust estimates and "best-guesses", neat.
Merging SpaceX with a public company like Tesla would create a lot of issues for the classified projects SpaceX does.
What sort of issues are you thinking?
Plenty of defense contractors with classified projects are already publicly listed, so this is not uncharted territory.
Lockhead Martin for example: https://investors.lockheedmartin.com/news-releases/news-rele...
Gives this level of detail:
> Aeronautics classified program losses $(950)
> MFC classified program losses -
It seems very safe from a national security perspective.
No? Almost every big defense contractor is publicly traded.
I imagine those are surmountable challenges. Boeing somehow manages.
But more likely that merger would consist of SpaceX acquiring Tesla and taking it private
There is no way Elon could raise the 1.4 trillion to take Tesla private
Raytheon is public.
As a SpaceX fan, I am saddened by this news.
The only reason for xAI to join SpaceX is to offload Elon's Twitter debt in the upcoming IPO.
This is just what I was thinking.
Twitter (X) owed $1.3B in debt every year in interest since Musk's takeover. This was before re-financing in a higher interest rate environment. The company was losing $200MM+ per year on ~$5B in revenue before the takeover, and there are reports that revenues have decreased by round 50%.
Best case scenario if we accept those numbers is that X makes $3B per year and about half of that goes immediately out the door in debt payments before paying a cent for the entire business to function.
However, if SpaceX acquires X, that ~$1.5B in interest is a fraction of the $8B In profits SpaceX is allegedly generating annually. Further, they can restructure the debt if it's SpaceX's debt, and not owned by X. Investors will be more likely to accept SpaceX shares as collateral than X.
> The company was losing $200MM+ per year on ~$5B in revenue before the takeover, and there are reports that revenues have decreased by round 50%.
X made a profit last year because they cut costs lower than the drop in ad revenue (which is also slowly recovering). The big question is if they will still be profitable in 2026 year without the US election driving big traffic numbers and ads.
Either this is a straight up con, or Musk found a glitch in physics. It's extremely difficult to keep things cold in space.
He buys twitter at an inflated valuation. Runs it to the ground to a much lower valuation of $9B. [1] Then, his company Xai buys Twitter at a $33B, inflating the valuation up. Then SpaceX merges with Xai for no particular reason, but is expected to IPO at a $1T+ in the upcoming years. [3]
I’m not that smart, but if I were, I would be thinking this is an extended way to move the losses from the Twitter purchase on to the public markets.
[1] https://www.axios.com/2023/12/31/elon-musks-x-fidelity-valua...
[2] https://www.reuters.com/markets/deals/musks-xai-buys-social-...
[3] https://www.cnbc.com/amp/2026/02/02/elon-musk-spacex-xai-ipo...
It also makes it impossible for Twitter/X to die, as it deserves. It is by far the most toxic mainstream social network. It has an overwhelming amount of far right supremacist content. So bad that it literally resulted in Vivek Ramaswamy, a gubernatorial candidate in Ohio, to quit Twitter/X - nearly 100% of replies to his posts were from far right racists.
Obviously advertisers have not been fans. And it is a dying business. But rather than it dying, Elon has found a clever (and probably illegal) way to make it so that SpaceX, which has national security importance, is going to prop up Twitter/X. Now our taxpayer dollars are paying for this outrageous social network to exist.
I find HN and the tech circles to be one of the main community pillars holding up X. None of my social friends use it anymore, but links absolutely abound here, and it seems like the standard line is to pretend Elon, Grok, all the one button revenge and child porn etc don’t exist. I truly can’t fathom the amount of not thinking about it it would take to keep using the platform.
Just use lists, "Your Followers" tab and never touch the "For You" tab and its basically the same as Twitter was 5 years ago.
I am with you 100%.
It was easy to support SpaceX, despite the racist/sexist/authoritarian views of its owner, because he kept that nonsense out of the conversation.
X is not the same. Elon is actively spewing his ultraconservative views on that site.
Now that these are the same company, there's no separation. SpaceX is part of Musk's political mission now. No matter how cool the tech, I cannot morally support this company, and I hope, for the sake of society, it fails.
This announcement, right after the reveal that Elon Musk reached out to Jeffrey Epstein and tried to book a trip to Little St. James so that he could party with "girls", really doesn't bode well.
It's a shame you can't vote these people out, because I loved places like Twitter, and businesses like SpaceX and Tesla, but Elon Musk is a fascist who uses his power and influence to attack some of the most important pillars of our society.
You kinda can, just don’t make a Twitter account, don’t buy teslas, don’t use grok. Tell your friends
Setting aside the possibility it's window dressing for a financial bailout, there would be two ways compute in space makes sense:
1) new technology improves vacuum heat radiation efficiency
2) new technology reduces waste heat generation from compute
All the takes I've seen have been focused on #1, but I'm starting to wonder about #2... Specifically spintronics and photonic chips.
If you solve 2, heat dissipation goes away on earth too, so what’s the advantage of space
I'm not the best person to make that case as I can only speculate (land cost, permitting, latency, etc). /Shrug
In all the conversations I've seen play out on hacker news about compute in space, what comes up every time is "it's unviable because cooling is so inefficient".
Which got me thinking, what if cooling needs dropped by orders of magnitude? Then I learned about photonic chips and spintronics.
> space is called “space” for a reason.
you think we don't have enough space on earth for a few buildings? this seems like a purely western cope. China seems perfectly able to build out large infrastructure projects with a land area smaller than that of the continentenal USA
> new technology improves vacuum heat radiation efficiency
Isn't this fixed by blackbody radiation equations?
This isn't really true, though? The ISS does it with radiators that are ~1/2 the area of its solar panels, and both should scale linearly with power?
ISS radiators run on water and ammonia. Think about how much a kg costs to lift to space and you'll see the economics of space data centers fall apart real fast. Plus, if the radiator springs a leak the satellite is scrap.
I don't pretend to understand the thermodynamics of all of this to do an actual calculation, but note that the ISS spends half its time in the shadow of the earth, which these satellites would not do.
The ISS creates radically less heat than a datacenter
Moving electricity long distance is a lot easier than moving coolant long distances, which puts a soft limit on the reasonable size of the solar array of these satellites. But as long as you stay below that and pick a reasonable orbit it's indeed not too bad, you just have to properly plan for it
The ISS isn't consuming and generating megawatts+ of power.
Yes but if the solar panel area scales linearly with radiator area, the problem doesn't get worse?
It does if you don't turn off the heat source every 30 minutes or so. Since the "datacenters" are targeted at sun synchronous orbits they have 24/7 heat issues. And they convert pretty much all collected energy into heat as well (and some data, but that's negligible). Those GPUs are not magically not generating heat.
Wouldn't the panels themselves need cooling too? The ones on earth generate heat while being in the sun.
There are commercial systems that can use open loop cooling (i.e. spray water) to improve efficiency of the panel by keeping the panel at a optimal temp of ~25C and the more expensive closed loop systems with active cooling recovers additional energy from the heat by circulating water like a solar heater in the panel back.
I would hope SpaceX is using more efficient solar cells than the ISS
Probably not. The ISS got a solar array upgrade after its initial launch:
https://www.spectrolab.com/company.html
Twenty-five years after the ISS began operations in low Earth orbit, a new generation of advanced solar cells from Spectrolab, twice as efficient as their predecessors, are supplementing the existing arrays to allow the ISS to continue to operate to 2030 and beyond. Eight new arrays, known as iROSAs (ISS Roll-Out Solar Arrays) are being installed on the ISS in orbit.
The new arrays use multi-junction compound semiconductor solar cells from Spectrolab. These cells cost something like 500 times as much per watt as modern silicon solar cells, and they only produce about 50% more power per unit area. On top of that, the materials that Spectrolab cells are made of are inherently rare. Anyone talking about scaling solar to terawatts has to rely on silicon or maybe perovskite materials (but those are still experimental).
It's a con, his AI business is failing, so he's rolling it up into the profitable business. Did a similar thing with Twitter.
This is so obvious, but it's so stupid and at this scale that people find it hard to believe.
Existing satellites manage to keep their equipment that already can consume several kW cool just fine.
You might need space for radiators, but there is plenty space in space.
5,000 Starship launches to match the solar/heat budget of the 10GW "Stargate" OpenAI datacenter. The Falcon 9 family has achieved over 600 launches.
The ISS power/heat budget is like 240,000 BTU/hr. That’s equivalent to half of an Nvidia GB200 NVL72 rack. So two international space stations per rack. Or about 160,000 international space stations to cool the 10GW “Stargate” datacenter that OpenAI’s building in Abilene. There are 10,000 starlink satellites.
Starship could probably carry 250-300 of the new V2 Mini satellites which are supposed to have a power/heat budget of like 8kW. That's how I got 5,000 Starship launches to match OpenAI’s datacenter.
Weight seems less of an issue than size. 83,000 NVL72’s would weigh 270 million lbs or 20% of the lift capacity of 5000 starship launches. Leaving 80% for the rest of the satellite mass, which seems perhaps reasonable.
Elon's napkin math is definitely off though, by over an order of magnitude. "a million tons per year of satellites generating 100 kW of compute power per ton" The NVL72's use 74kW per ton. But that's just the compute, without including the rest of the fucking satellite (solar panels and radiators). So that estimate is complete garbage.
One note: If you could afford to send up one of your own personal satellites, it would be extremely difficult for the FBI to raid.
Several kW is nothing for a bank of GPUs.
Radiators in space are extremely inefficient because there's no conduction.
Also you have huge heat inputs from the sun. So you need substantial cooling before you get around to actually cooling the GPUs.
you put the radiators and the rest of the satellite within the shade of the solar panels, you can still make the area arbitrarily large
EDIT: people continue downvoting and replying with irrelevant retorts, so I'll add in some calculations
Let's assume
1. cheap 18% efficient solar panels (though much better can be achieved with multijunction and quantum-cutting phosphors)
2. simplistic 1360 W/m^2 sunlight orthogonal to the sun
3. an abstract input Area Ain of solar panels (pretend its a square area: Ain = L ^ 2)
4. The amount of heat generated on the solar panels (100%-18%) * Ain * 1360 W / m ^ 2, the electrical energy being 18% * Ain * 1360 W / m ^ 2. The electrical energy will ultimately be converted to computational results and heat by the satellite compute. So the radiative cooling (only option in space) must dissipate 100% of the incoming solar energy: the 1360 W / m^2 * Ain.
5. Lets make a pyramid with the square solar panel as a base, with the apex pointing away from the sun, we make sure the surface has high emissivity (roughly 1) in thermal infrared. Observe that such a pyramid has all sides in the shade of the sun. But it is low earth orbit so lets assume warm earth is occupying one hemisphere and we have to put thermal IR reflectors on the 2 pyramid sides facing earth, so the other 2 pyramid sides face actual cold space.
6. The area for a square based symmetric pyramid: we have
6.a. The area of the base Ain = L * L.
6.b. The area of the 4 sides 2 * L * sqrt( L ^ 2 / 4 + h ^ 2 )
6.c. The area of just 2 sides having output Area Aout = L * sqrt( L ^ 2 / 4 + h ^ 2 )
7. The 2 radiative sides not seeing the sun and not seeing the earth together have the area in 6.c and must dissipate L ^ 2 * 1360 W / m ^ 2 .
8. Hello Stefan-Boltzmann Law: for emissivity 1 we have the radiant exitance M = sigma * T ^ 4 (units W / m ^ 2 )
9. The total power exited through the 2 thermal radiating sides of the pyramid is then Aout * M
10. Select a desired temperature and solve for h / L (to stay dimensionless and get the ratio of the pyramid height to its base side length), lets run the satellite at 300 K = ~26 deg C just as an example.
11. If you solve this for h / L we get: h / L = sqrt( ( 1360 W / m ^ 2 / (sigma * T ^ 4 ) ) ^ 2 - 1/4 )
12. Numerically for 300K target temperature we get: h/L = sqrt((1360 / (5.67 * 10^-8 * 300 ^ 4)) ^ 2 - 1/4) = 2.91870351609271066729
13. So the pyramid height of "horribly poor cooling capability in space" would be a shocking 3 times the side length of the square solar panel array.
As a child I was obsessed with computer technology, and this will resonate with many of you: computer science is the poor man's science, as soon as a computer becomes available in the household, some children autodidactically educate themselves in programming etc. This is HN, a lot of programmers who followed the poor man's science path out of necessity. I had the opportunity to choose something else, I chose physics. No amount of programming and acquiring titles of software "engineer" will be a good substitute for physicists and engineers that actually had courses on the physical sciences, and the mathematics to follow the important historical deductions... It's very hard to explain this to the people who followed the path I had almost taken. And they downvote me because they didn't have the opportunity, courage or stamina to take the path I took, and so they blindly copy paste each others doomscrolled arguments.
Look I'm not an elon fanboy... but when I read people arguing that cooling considerations excludes this future, while I know you can set the temperature arbitrarily low but not below background temperature of the universe 4 K, then I simply explain that obviously the area can be made arbitrarily large, so the temperature can be chosen by the system designer. But hey the HN crowd prefers the layers of libraries and abstractions and made themselves an emulation of an emulation of an emulation of a pre-agreed reality as documented in datasheets and manuals, and is ultimately so removed from reality based communities like physics and physics engineering, that the "democracy" programmers opinions dominate...
So go ahead and give me some more downvotes ;)
It's really not that simple. See this for a good explanation of why: https://taranis.ie/datacenters-in-space-are-a-terrible-horri...
It all basically boils down to: in order to dissipate heat, you need something to dissipate heat into, e.g. air, liquid, etc. Even if you liquid cool the GPUs, where is the heat going to go?
On Earth, you can vent the heat into the atmosphere no problem, but in space, there's no atmosphere to vent to, so dissipating heat becomes a very, very difficult problem to solve. You can use radiators to an extent, but again, because no atmosphere, they're orders of magnitude less effective in space. So any kind of cooling array would have to be huge, and you'd also have to find some way to shade them, because you still have to deal with heat and other kinds of radiation coming from the Sun.
It's easier to just keep them on Earth.
That helps with the heat from the sun problem, but not the radiation of heat from the GPUs. Those radiators would need to be unshaded by the solar panels, and would need to be enormous. Cooling stuff in atmosphere is far easier than in vacuum.
Not so. Look at the construction of JWST. One side is "hot", the other side is very, very cold.
I am highly skeptical about data centers in space, but radiators don't need to be unshaded. In fact, they benefit from the shade. This is also being done on the ISS.
That's fair. I meant they would need a clear path to open space not blocked by solar panels, but yes, a hot and cold side makes sense.
The whole concept is still insane though, fwiw.
Look at how many layers of insulation are needed for the JWST to have a hot and cold side! Again, this is not particularly simple stuff.
The JWST operates at 2kw max. That's not enough for a single H200.
AI datacenters in space are a non-starter. Anyone arguing otherwise doesn't understand basic thermodynamics.
> arbitrarily large
Space is not empty. Satellites have to be boosted all the time because of drag. Massive panels would only worsen that. Once you boosters are empty the satellite is toast.
"Satellites have to be boosted all the time because of drag."
On Low Earth Orbits (LEOs), sure, but the traces of atmosphere that cause the drag disappear quite fast with increasing altitude. At 1000 km, you will stay up for decades.
I’ve got a perpetual motion machine to sell you.
arbitrarily large means like measured in square km. Starcloud is talking about 4km x 4km area of solar panels and radiative cooling. (https://blogs.nvidia.com/blog/starcloud/)
Building this is definitely not trivial and not easy to make arbitrarily large.
I'm not big on this subject, but I understand that heat transfer is difficult in space, because there's little to transfer to. If the solution is just making large radiators, then that means you're sending some big payloads full of radiators. Not to mention all the solar panels needed. I wanna live in sci-fi land too, but I don't see how it makes any sense compared to a terrestrial data center.
the radiators would be lighter compared to the solar panels, and slightly smaller surface area so you can line them back to back
If someone has a design out there where this works and you can launch it economically on a rocket today, I wanna see that. And then I wanna compare it to the cost of setting up some data centers on earth (which BTW, you can service in real time, it sounds like these will be one-and-done launches).
> keep their equipment that already can consume several kW cool just fine
That's equivalent to a couple datacenter GPUs.
> You might need space for radiators, but there is plenty space in space.
Finding space in space is the least difficult problem. Getting it up there is not easy.
You can line the solar panels and radiators facing away from each other, and the radiators would take up less surface area. I think maybe the tricky part would be the weight of water + pipes to move heat from the compute to the radiators.
Water is not needed to move heat. Heat pipes do it just fine. There's one in your laptop and one in your phone too. It does scale up.
Interesting, That could surely simplify things.
There's plenty of space in space, but there isn't plenty of space in rocket fairings, nor is there plenty of lift capacity for an unlimited amount of radiators.
Maybe Karpathy has been hired to design a Full Self Cooling system.
> It's extremely difficult to keep things cold in space.
This is one of those things that's not obvious till you think about it.
It's such bullshit that we've decided this moron and others in his cohort can unilaterally reallocate such vast portions of humanity's labor at their whims.
This is an extremely stupid idea, but because of our shared delusion of capitalism and the idea that wealth accumulation at the top should be effectively limitless, this guy gets to screw around and divert actual human labor towards insane and useless projects like this rather than solving real world problems.
Just put a fan in a window.
what makes you believe this?
radiators can be made as long as desirable within the shade of the solar panels, hence the designer can pracitically set arbitrarily low temperatures above the background temperature of the universe.
Radiators can shadow each other, so that puts some kind of limit on the size of the individual satellite (which limits the size of training run it can be used for, but I guess the goal for these is mostly inference anyway). More seriously, heat conduction is an issue: If the radiator is too long, heat won't get from its base to its tip fast enough. Using fluid is possible, but adds another system that can fail. If nothing else, increasing the size of the radiator means more mass that needs to be launched into space.
Shading does work; JWST does this. However I don't see how you can make it work for satellite data centers. You would constantly be engaging attitude control as you realigned the panels to keep the radiators in shade. You'd run out of thruster fuel so fast you'd get like a month out of each satellite
https://en.wikipedia.org/wiki/Spacecraft_attitude_determinat...
Radiators can only be made as long as desirable because there's gravity for the fluid inside to go back down once it condenses. Even seen those copper heat pipes in your PC radiator?
Fluid in heat pipes moves through capillary action.
these same comments pop up every time someone brings up satellite data-centers where people just assume the only way of dissipating heat is through convection with the environment.
No, we just "assume" (i.e. know) that radiation in a vacuum is a really bad way of dissipating heat, to the point that we use vacuum as a very effective insulator on earth.
Yes, you can overcome this with enough radiator area. Which costs money, and adds weight and space, which costs more money.
Nobody is saying the idea of data centers in space is impossible. It's obviously very possible. But it doesn't make even the slightest bit of economic sense. Everything gets way, way harder and there's no upside.
Additional radiator area means bigger spacecraft, implies more challenge with attitude control. Lower down you get more drag so you use propellant to keep yourself up, higher up you have more debris and the large area means you need to frequently manoeuvre to avoid collisions. Making things bigger in space is not trivial! You can't just deploy arbitrarily large panels and expect everything to be fine.
The radiators would be lighter compared to the solar panels, and slightly smaller surface area so you can line them back to back
I don't think dissipating heat would be an issue at all. The cost of launch I think is the main bottleneck, but cooling would just be a small overhead on the cost of energy. Not a fundamental problem.
The pertinent thing is that it’s not an advantage. It may be doable but it’s not easier than cooling a computer in a building.
what? the heat is coming from inside the house
The Goalpost shift continues, If elon were working for me, I would have fired him for having never delivered on any of his projects.
Hyperloop > Neuralink > Self-Driving Cars > Robotaxi fleets > Personal Robots > Orbital Datacenters > [Insert next]
> If elon were working for me, I would have fired him for having never delivered on any of his projects.
Never? For the sheer amount of moonshot bets he's doing, his track record would make any VC jealous. Zip2, PayPal, Tesla, SpaceX, Grok/xAI.
World's Best At Surfing A Temporary Hyperinflation Wave is not a life goal to really be proud of tbh.
>Zip2
I guess props to scamming Compaq into making a large investment that didn't pan out. He did personally make money so I guess win for him.
>In an effort to woo investors, Elon Musk built a large casing around a standard computer to give the impression that Zip2 was powered by a supercomputer.
>PayPal
Huh? He didn't found Paypal, his company was acquired by Paypal. You might as well give him credit for eBay while you're at it. Paypal released their first digital wallet in 1999. They acquired x.com (and Musk) in 2000. Paypal itself was then acquired by eBay in 2002.
>Tesla
Investor, not founder.
>SpaceX
Yup, props here.
>Grok/xAI
Hasn't made a penny, no signs it had any path to profitability, which is why it was shoved into Space-X to cover his personal losses.
Hey now. Huperliop was designed to scuttle California’s light rail project. Which it did. Mission accomplished.
I think you mean "California High-Speed Rail", not light rail.
Light rail, generally refers to urban rail, "trams".
What are you talking about?
This is the funniest conspiracy theory.
It’s not really so much a conspiracy theory as a thing that he outright said.
https://www.jalopnik.com/did-musk-propose-hyperloop-to-stop-...
That link denies the conspiracy theory?
> “Or did he just have an idea and blurt it out," I asked Vance. > "I'm 99.9-percent sure it's the latter," Vance tells me.
Also that to scapegoat Musk for killing the California train when California was perfectly able to kill it itself:
> Vance then brought up a valid point: "In all this time we've been talking about high-speed rail, there's still almost none that's built....
hmm Tesla shipped millions of cars SpaceX launches 90% of space payloads, Starlink is working well. Thats hard to categorize as never delivered on any of his projects
The crucial thing is that Tesla's valuation has the hype projects baked in. The fact that it never delivered self driving or a robotaxi fleet and is now being saved solely by an import ban on Chinese EVs means that any success he had with Tesla is now an illusion.
I don't follow his promises but have seen first hand how far ahead Tesla FSD is compared to competitors in the consumer space. It's not even close.
This current announcement seems silly, though.
Yep but again that does not qualify as never delivered on anything.
> The fact that it never delivered self driving or a robotaxi fleet a
Once again pointing out Tesla has around 300 robotaxis running in 2 cities (Austin/SF).
The projects promised to be life altering for all mankind, they ended up being not even life altering for super rich Americans considering that Teslas are just EVs which without FSD are just regular cars with a different propellent that were made for political purposes and virtue signaling
The EV revolution has always been something almost dystopic : Trillions of dollars spent in order to not have the slightest amount of quality of life improvement, if anything a worse quality of life because you buy an EV that you cannot use 24/7/365 whereas you can an ICE car for much less .
As soon as something kinda elegant and charged with hope as far as collective quality of life improvement is concerned (AI/ChatGPT) the whole green/EV revolution rightfully went out the window
If Musk was this genius you guys make him to be at 50 and with all the capital he burned he should have at least one company that if you disappeared the world would look drastically different, like if you disappeared Microsoft or Apple or Exxon or Aramco the world would come to a screeching halt.
Disappear one of Musk companies and everything would be the same as he's always involved in these sort of aspirational companies which have this great vision always 5 years into the future that never materialize into anything tangible
well Tesla did jump start the EV revolution not life altering but is pretty important. IF SpaceX gets spaceship right that will be a huge leap forward.
Sir, your comment appears to qualify as "moving the goal post". TSLA never delivered a single inexpensive electric vehicle, and just last week abandoned all high-end efforts (S/X/CT discontinued). All TSLA manufactures now are overpriced "meh" transport boxes. Yes, TSLA was early, and now they are far, far behind the competition.
Can we evaluate based on the stated goals, or why does the criteria keep shifting?
Yeah, but there are enough people to buy the hype.
In a way, its kind of cool to see how robber barons work in real time in our generation. Its also insanely depressing as they will systematically enshittify and extract as much wealth from society as is possible.
I don't actually think the Robber Barons in the 1920s had people going out of their way to defend them and insist they had special knowledge.
The New Deal happened with massive popular support because people did not like the Barons, and wanted to stop them and actually have a life worth living.
It only took like 30 years of suffering.
> "The only logical solution therefore is to transport these resource-intensive efforts to a location with vast power and space. I mean, space is called 'space' for a reason. [crying laughing emoji]"
This is all the reasoning provided. It is quite sad how a company I admired so much has become embroiled in financial doohickery.
> The basic math is that launching a million tons per year of satellites generating 100 kW of compute power per ton would add 100 gigawatts of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 1 TW/year from Earth.
> My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space.
This is so obviously false. For one thing, in what fantasy world would the ongoing operational and maintenance needs be 0?
You operate them like Microsoft's submerged data center project: you don't do maintenance, whatever fails fails. You start with enough redundancy in critical components like power and networking and accept that compute resources will slowly decrease as nodes fail
No operational needs is obviously ... simplified. You still need to manage downlink capacity, station keeping, collision avoidance, etc. But for a large constellation the per-satellite cost of that would be pretty small.
How do you make a small fortune? Start with a big one.
The thing being called obvious here is that the maintenance you have to do on earth is vastly cheaper than the overspeccing you need to do in space (otherwise we would overspec on earth). That's before even considering the harsh radiation environment and the incredible cost to put even a single pound into low earth orbit.
If you think the primary source of electricity is solar (which clearly Musk does), then space increases the amount of compute per solar cell by ~5x, and eliminates the relatively large battery required for 24/7 operation. The thermal radiators and radiation effects are manageable.
The basic idea of putting compute in space to avoid inefficient power beaming goes back to NASA in the 60s, but the problem was always the high cost to orbit. Clearly Musk expects Starship will change that.
My dude, ISS has 200 KW of peak power.
NVIDIA H200 is 0.7 KW per chip.
To have 100K of GPUs you need 500 ISSs.
ISS cooling is 16KW dissipation. So like 16 H200. Now imagine you want to cool 100k instead of 16.
And all this before we talk about radiation, connectivity (good luck with 100gbps rack-to-rack we have on earth), and what have you.
—
Sometimes I think all this space datacenters talk is just a PR to hush those sad folks that happen to live near the (future) datacenter: “don’t worry, it’s temporary”
https://www.nytimes.com/2025/10/20/technology/ai-data-center...
The ISS is in the middle of rolling out upgrades to their panels so it’s not a great comparison. https://en.wikipedia.org/wiki/Roll_Out_Solar_Array
> ROSA is 20 percent lighter (with a mass of 325 kg (717 lb))[3] and one-fourth the volume of rigid panel arrays with the same performance.
And that’s not the current cutting edge in solar panels either. A company can take more risks with technology choices and iterate faster (get current state-of-the-art solar to be usable in space).
The bet they’re making is on their own engineering progress, like they did with rockets, not on sticking together pieces used on the ISS today.
Now tell me how you heat dissipate all this. Not that there is a lot of air or water in space.
Not that you would want 500+ square meters just for cooling of 200KW
And, mind you, it won’t be a simple copper radiator
https://www.nasa.gov/wp-content/uploads/2021/02/473486main_i...
The idea here is that the economics of launch are changing with Starship such that the "incredible cost" and "overspeccing" of space will become much less relevant. There's a world where, because the cost per kg is so low, a data center satellite's compute payload is just the same hardware you'd put in a terrestrial rack, and the satellite bus itself is mass-produced to not-particularly-challenging specs. And they don't have to last 30 years, just 4-ish, when the computer is ready for retirement anyway.
Will that come to be? I'm skeptical, especially within the next several years. Starship would have to perform perfectly, and a lot of other assumptions hold, to make this make sense. But that's the idea.
How much maintenance do you need? Lets say you have hardware whose useful lifespan due to obsolescence is 5 years, and in 4, the satellite will crash into the atmosphere anyways.
Let's say given component failure rates, you can expect for 20% of the GPUs to fail in that time. I'd say that's acceptable.
Radiation is a bitch. Especially at those nanometers and memory bandwidth.
And cooling. There is no cold water or air in space.
If you ramp up the economies of scale to make those things - radiation protection and cost per pound - the calculus changes. It's supposed to synergize with Starship, and immediately take advantage of the reduced cost per pound.
If the cost per pound, power, regulatory burden, networking, and radiation shielding can be gamed out, as well as the thousand other technically difficult and probably expensive problems that can crop up, they have to sum to less than the effective cost of running that same datacenter here on earth. It's interesting that it doesn't play into Jevon's paradox the way it might otherwise - there's a reduction in power consumption planetside, if compute gets moved to space, but no equivalent expansion since the resource isn't transferable.
I think some sort of space junk recycling would be necessary, especially at the terawatt scale being proposed - at some point vaporizing a bunch of arbitrary high temperature chemistry in the upper atmosphere isn't likely to be conducive to human well-being. Copper and aluminum and gold and so on are also probably worth recovering over allowing to be vaporized. With that much infrastructure in space, you start looking at recycling, manufacturing, collection in order to do cost reductions, so maybe part of the intent is to push into off-planet manufacturing and resource logistics?
The whole thing's fascinating - if it works, that's a lot of compute. If it doesn't work, that's a lot of very expensive compute and shooting stars.
some people really gotta stop huffing VC fumes
Or, just saying, be critical of ideas and think them through, and take in what experts say about it, and determine for yourself what's up. If a bunch of people who usually seem to know what they're talking about think there's a legitimate shot at something you, as a fellow armchair analyst, think is completely impractical, it makes sense to go and see if maybe they know something you don't.
In this case, it's all about Starship ramping up to such a scale that the cost per pound to orbit drops sufficiently for everything else to make sense - from the people who think the numbers can work, that means somewhere between $20 and $80 per pound, currently at $1300-1400 per pound with Falcon 9. Starship at scale would have to enable at least 2 full orders of magnitude decrease in price to make space compute viable.
If Starship realistically gets into the $90/lb or lower range, space compute makes sense; things like shielding and the rest become pragmatic engineering problems that can be solved. If the cost goes above $100 or so, it doesn't matter how the rest of the considerations play out, you're launching at a loss. That still might warrant government, military, and research applications for space based datacenters, especially in developing the practical engineering, but Starship needs to work, and there needs to be a ton of them for the datacenter-in-space idea to work out.
Note how Musk cleverly doesn't claim that not doing maintenance drives down costs.
Nothing in there is a lie, but any substance is at best implied. Yes, 1,000,000 tons/year * 100kW/ton is 100GW. Yes, there would be no maintenance and negligible operational cost. Yes, there is some path to launching 1TW/year (whether that path is realistic isn't mentioned, neither what a realistic timeline would be). And then without providing any rationale Elon states his estimate that the cheapest way to do AI compute will be in space in a couple years. Elon is famously bad at estimating, so we can also assume that this is his honest belief. That makes a chain of obviously true statements (or close to true, in the case of operating costs), but none of them actually tell us that this will be cheap or economically attractive. And all of them are complete non-sequiturs.
Elon might have a scoop on getting things to orbit cheaper than everyone else.
Unless I missed something the Microsoft underwater data center was basically a publicity stunt.
Anyone who thinks it makes sense to blast data centers into space has never seen how big and heavy they are, or thought about their immense power consumption, much less the challenge of radiating away that much waste heat into space.
Radiation is an even bigger problem, especially in the polar orbits they are talking about.
It’s only a problem if you get the machines up there! Which I’d argue is economically unviable to boot.
I don't think it was a stunt. It was an experiment.
I think passive cooling (running hot) reduced some of the advantages of undersea compute.
Ironically a benefit of underwater datacenters would be reduced cosmic rays. Not so great in orbit, I imagine!
I was listening to a Darknet Diaries episode where Maxie Reynolds seems to make it work: https://subseacloud.com/ I don't know how profitable they are, and I doubt this is scalable enough, but it can work as a business.
What about a data centre only running SQLite?
Well the thing is that it seemed to have been successful beyond all expectations despite being that? They had fewer failures due to the controlled atmosphere, great cooling that took no extra power, and low latency due to being close to offshore backbones. And I presume you don't really need to pay for the land you're using cause it's not really on land. Can one buy water?
Space is pretty ridicolous, but underwater might genuinely be a good fit in certain areas.
Hot saltwater is the worst substance on earth, excepting, maybe, hydrofluoric acid. You really don't want to cool things with ocean water over an extended period of time. And filtering/purifying it takes vast amounts of power (e.g. reverse osmosis).
My 4 Cylinder Diesel Volvo Penta is cooled by sea water. There is an elbow that may have to be replaced every few years,
I thought they had an issue with stuff growing on the cooling grates. Life likes to find warm water.
An 8 GPU B200 cluster goes for about $500k right now. You'd need to put thousands of those into space to mimic a ground-based data center. And the launch costs are best case around 10x the cost of the cluster itself.
Letting them burn up in the atmosphere every time there's an issue does not sound sustainable.
A Falcon Heavy takes about 63 tons to LEO, at a cost of about $1,500 per kg. A server with 4x H200s and some RAM and CPU costs about $200k, and weighs about 60kg, with all the cooling gear and thick metal. As is, it would cost $90k to get to LEO, half of the cost of the hardware itself.
I suppose that an orbit-ready server is going to cost more, and weigh less.
The water that serves as the coolant will weigh a lot though, but it can double as a radiation shield, and partly as reaction mass for orbital correction and deorbiting.
Are launch costs really 10x!? Could I get a source for that?
In the back on my head this all seemed astronomically far-fetched, but 5.5 million to get 8 GPUs in space... wild. That isn't even a single TB of VRAM.
Are you maybe factoring in the cost to powering them in space in that 5 million?
I guess he adds the weight of all the hardware to make the whole thing work.
You also need square kms of radiators to cool 100MW
Playing devil's advocate, when a GPU dies you don't typically fix it, right? You just replace it.
What if you could keep them in space long enough that by the time they burn up in the atmosphere, there are newer and better GPUs anyway?
Still doesn't seem sustainable to me given launch costs and stuff (hence devil's advocate), but I can sort of see the case if I squint?
You mean you operate them like Microsoft's failed submerged data center project [1]. When pointing at validating past examples you are generally supposed to point at successes.
[1] https://en.wikipedia.org/wiki/Project_Natick
The opposite of down is up, so it wouldn't be completely illogical.
Did we read the same Wikipedia page? It doesn't say the word "failed" anywhere on it.
> By 2024, Project Natick had been inactive for several years, though it was referenced in media as though it was ongoing. That year, Microsoft confirmed that the project was inactive and that it had no servers underwater.
I wouldn't exactly call this a success, for that matter.
But if we’re going down that line of thinking then it’s a poor comparison. I could open a data centre on the ground and use the same principle of zero maintenance, and it would be way cheaper and way more powerful.
How many submerged datacenters is Microsoft operating?
None, https://www.datacenterdynamics.com/en/news/microsoft-confirm...
Zero.
But you could just run a “zero maintenance” data center on Earth and not pay to blast it into orbit.
You also have to get rid of waste heat :P
This will totally work since we have an unlimited amount of rare earth elements we can just ship off into space never to see again. Infinite raw materials + infinite power equals infinite AI!!!
Accountants love this
The financial system is already freaking out about depreciation on land based data centers. I don't think it could survive what you're talking about.
Being under the ocean in a metal box you don't get too many micro-meteors or cosmic rays though.
...and costs pennies compared to putting anything up there, so it can even enjoy those cosmic goodies.
Just like you don’t get much water in space.
My understanding was that access to very large body of cold water was a core feature for the project. The water was to be used for cooling relatively efficiently or cheaply.
As soon as a statement contains a timeframe estimate by Musk you know to disregard it entirely.
The thing is: at the end of the day, SpaceX takes the "impossible" and makes it "late".
People are going to Tory Bruno the space datacenters until one day their Claude agent swarm's gonna run in space and they'll be wondering "how did we get here"?
The thing is: at the end of the day, making absolute statements about the inevitability of future success is a fool’s errand.
Musk has a documented history of failing to deliver on promises, timescale or no. So it’s best to engage in some actual critical thinking about the claims he is making.
napkin math says sq kms of radiators to cool 100MW, it's just patently ridiculous
Or takes the impossible and puts a half baked version of it behind a $99/ month paywall.
Pay wallet? Starlink was never gonna be free.
There's clearly rhetorical hyperbole happening there. But assuming that thermal rejection is good in space, & launch costs continue falling, as earth-based data centers become power/grid-constrained, there is a viable path for space power gen.
The craziest part of those statements is "100 kW per ton." IDK what math he is doing there or future assumptions, but today we can't even sniff at 10 kW per ton. iROSA [1] on the ISS is about 0.150 kW per ton.
[1]https://en.wikipedia.org/wiki/Roll_Out_Solar_Array
edit: iROSA = 33 kW per ton, thanks friends
Not to be an Elon defender, but can you back up your 0.15/ton? My own searching puts ROSA orders of magnitude higher. Each array is 600kg (0.6t) and puts out 20kw (https://space.skyrocket.de/doc_sdat/irosa-1.htm) which makes 20/0.6 = 33.333 kw/ton
You're right, my fault. I made an math booboo somewhere. Your calc seems right
Hey all good. My advice, not that you asked for it, is to put the math in the comment. Even as a footnote. I've found myself backtracking a lot of math comments after I stare at it in the text box for a few seconds.
The company lists their ISS solar panels as 28 kW for 331 kg, which comes pretty near to 100 W/kg.
Company website:
https://rdw.com/wp-content/uploads/2023/06/redwire-roll-out-...
And their Opal configuration beats the metric: 5.3 kW for 42.7 kg.
Maybe I'm just out of the loop, but is solar substantially more efficient in space? I assume the satellites won't orbit in a way that follows the sun. And presumably the arrays of panels they can attach to a satellite don't exceed the size of the panels you could slap on and around a data center (at least without being insanely expensive).
Yeah the main benefits are:
1. solar is very efficient at generating energy, no moving parts, simple physics etc.
2. in space you don't deal with weather or daylight cycle, you can just point your panels at the sun and generate very stable energy, no batteries required
3. environmental factors are simpler, no earthquakes, security, weather. Main problem here is radiation
In theory its a very elegant way to convert energy to compute.
2 is wrong. At a Lagrange point you can do this. Not in low earth orbit - in LEO sunset is every 60 minutes or so, and you spend the next 60 minutes in darkness.
Satellites are heavily reliant on either batteries or being robust to reboots, because they actually do not get stable power - it's much more dynamic (just more predictable too since no weather).
Interesting.
According to this other source https://www.satellitetoday.com/connectivity/2026/02/02/space...
the filing mentions this
> these satellites would operate between 500 km and 2,000 km altitude and 30 degrees and Sun-Synchronous Orbit inclinations (SSO)
> But assuming that thermal rejection is good in space
Don't assume this. Why would you assume this?
Just put a slightly larger solar array on the same equipment on earth?
Solar in space is a very different energy source in terms of required infrastructure. You don't need batteries, the efficiency is much higher, cooling scales with surface area (radiative cooling doesn't work as well through an atmosphere vs. vacuum), no weather/day cycles. Its a very elegant idea if someone can get it working.
Only if you also disregard all the negatives.
The panels suffer radiation damage they don't suffer on Earth. If this is e.g. the same altitude orbits as Starlink, then the satellites they're attached to burn up after around tenth of their ground-rated lifetimes. If they're a little higher, then they're in the Van Allen belts and have a much higher radiation dose. If they're a lot higher, the energy cost to launch is way more.
If you could build any of this on the moon, that would be great; right now, I've heard of no detailed plans to do more with moon rock than use it as aggregate for something else, which means everyone is about as far from making either a PV or compute factory out of moon rock as the residents of North Sentinel Island are.
OK, perhaps that's a little unfair, we do actually know what the moon is made of and they don't, but it's a really big research project just to figure out how to make anything there right now, let alone making a factory that could make them cost-competitive with launching from Earth despite the huge cost of launching from Earth.
> put a slightly larger solar array on the same equipment on earth?
Land and permitting. I’m not saying the math works. Just that there are envelopes for it to.
There is practically infinite land in which to build a datacenter.
The math literally works.
The US mandates by law that we grow a fuck ton of corn to mix 10% ethanol into gasoline.
If you replaced just those cornfields with solar/wind, they would power the entire USA and a 100% electric vehicle fleet. That includes the fact that they are in the corn belt with less than ideal sun conditions.
We aren’t even talking about any farmland that produces actual food or necessary goods, just ethanol as a farm subsidy program.
The US is already horrendously bad at land use. There’s plenty of land. There’s plenty of ability to build more grid capacity.
Hello fello Technology Connections watcher?
You know it!
Sun-synchronous orbit means there's no nightime for satellites in that orbit.
"There's clearly rhetorical hyperbole happening there" in a business paper is called lying
Name a unicorn whose early round pitch decks are 100% free of wishful thinking
Elon is a pathological liar and it’s crazy that he still gets sanewashed after all he’s done. It’s insanity that he hasn’t been kicked out of leading his companies, and it’s also insanity that he hasn’t been prosecuted by the SEC.
You’ve spent too much life force trying to even understand the liar’s fake logic.
Let’s start right here: there is no such thing as becoming power/grid constrained on earth. If you replaced just the cornfields that the United States uses just to grow corn for ethanol in gasoline just in the corn belt, you could power the entire country with solar+batteries+wind. Easily, and cheaply.
If you don’t even believe that solar+batteries are cheap (they are), fine, choose your choice of power plant. Nuclear works fine.
The truth is, xAI combining with SpaceX is almost certainly corrupt financial engineering. SpaceX as a government contractor and that means Elon’s pal Trump can now siphon money into xAI via the federal government.
This is par for the course for an Elon-associated endeavor but it's been leaking out into the broader tech sector; make ludicrous claims and promises and somehow investors just throw money at you. FSD has been around the corner for over a decade, martian colonization will be here by the end of the decade for the past 20 years and General SuperAI will be here in a few years for the past few years.
It's always 2-3 years with this guy
Conveniently, about the amount of time it takes for the average person to forget and/or rematerialize in a new parallel dimension
The average person is not aware or does not care about Elon Musk’s claims and whether or not they come true.
You’re just not going at the speed of light as this guy’s brain is, time dilation is a thing
Currently, just a cursory google search shows $1500-3000 per kilogram to put something into low earth orbit. Lets take the low bound because of efficiencies of scale. So $1500.
A million tons will cost $1500x1000x1000000= 1,500,000,000,000. That is one and a half TRILLION dollars per year. That is only the lift costs, it does not take into account the cost of manufacturing the actual space data centers. Who is going to pay this?
That launch cost is remarkably cheap to someone that's handled a $1.5million dollar 5U server filled with GPUs and RAM that weighs under 100kg.
Obviously the solar and cooling for the above would both weigh and cost a ton but... It's feels surprisingly close to being within an order of magnitude of current costs when you ballpark it?
Like i don't think it's actually viable, it's just a little shocking that the idea isn't as far out of line as i expected.
That's the price before Starship which would be the prerequisite for this whole project.
Yes, and as we know Starship will be doing regular commercial launches starting in 2020, maybe 2021.
We're getting close to having the time for Starship's delays to be the same as the actual time for the Saturn 5 to go from plans to manned launches (Jan 1962-Dec 1968).
Are you trying to say it'll be delayed or that it'll never work?
One is obviously true, and the other is very likely false.
It’s hard to estimate what Starship’s actual costs will be when it isn’t fully operational. I am finding estimates of $100 to $200 per kilogram and even as low as $10 per kilogram.
Let’s say the costs in 5 years do get as low as $15 per kilogram or about 2 orders of magnitude improvement in launch prices. That means a 200-ton payload Starship would cost $3,000 to launch.
Do you honestly believe that? The world’s largest rocket cost a total of $3,000 to launch?
> Let’s say the costs in 5 years do get as low as $15 per kilogram or about 2 orders of magnitude improvement in launch prices. That means a 200-ton payload Starship would cost $3,000 to launch. Do you honestly believe that? The world’s largest rocket cost a total of $3,000 to launch?
You have missed three zeroes in this calculation ;)
15 per kg for a 200-ton payload is about 3 million$. That seems achievable, given that propellant costs are about 1-1.5 million.
Ah yeah. 200 tons is 200,000 kilograms. Definitely way off there. That is an incredulous number.
"it'll never work" is quite black and white while "failure" is a lot more of a grey area. Will it actually launch? Sure, we've seen it. Will it actually hit the reliability as sold? Will it have as fast of turnaround time to reach launch timing goals? Can it actually launch as much payload as promised? Will the economics actually shake out as intended?
Did the Cybertruck "never work"? Obviously not, they're on the streets. Was it a <$40k truck with >250mi range? No.
Did FSD "never work"? Obviously not, tons of people drive many, many miles without touching the wheel. Does Tesla feel confident in it enough to not require safety operators to follow it on robotaxi trips? No. Does Tesla trust it enough to operate in the Las Vegas Loop? No. Has Tesla managed to get any state to allow it to operate truly autonomously? No.
Look, I hope Starship does work as advertised. Its cool stuff. But I don't see it as a given that it will. And given by the track record of the guy who promised it, it gives even less confidence. I'm sad there's less competition in this space. We have so many billionaires out there and yet so few out there actually willing push envelopes.
One reliable method of pushing envelopes, attracting investment and hiring smart people is to get excited about unrealistic timelines.
The best case is you meed the unrealistic timeline, the average case outcome is you solve the problem but it is delayed several years. And the worst case is it fails and investors lose some money.
If you try to hire people but your message is: we want to reduce the cost of access to space by 20% in thirty years, you are going to get approximately zero competent engineers, and a whole lot of coasters.
And no investors, so you'll be dependent on the government anyway. Depending on the government is great until people you do not agree with or are generally anti science, are in power. I assume this part should not need an example nowadays?
> One reliable method of pushing envelopes, attracting investment and hiring smart people is to get excited about unrealistic timelines.
Its also a good way to shred morale and investor confidence when you're a decade past your timelines or continue to fail on actually delivering on past promises.
You'd think so, but if you bet on this guy not being able to get investors you'll end up being wrong.
It doesn't make sense (neither does Tesla's valuation, for example), but it is what it is.
Both Spacex and Xai have investors lining up.
The ISS’s solar arrays each weigh a metric ton and generate 35 KW a piece[0], and that’s just for the power collection.
They’d need incredible leaps in efficiency for an orbiting ton collecting and performing 100 KW of compute.
[0] https://en.wikipedia.org/wiki/Electrical_system_of_the_Inter...
> in what fantasy world would the ongoing operational and maintenance needs be 0?
Well, if you can't get there, you can't do maintenance, so there is zero maintenance :)
Satellites have large operational costs. Satellite fleets even more so.
I remember reading somewhere that satellites are extra expensive for 2 reasons:
- launch costs are so high that doing exotic bespoke engineering might be worth it if it can shave off a few pounds
- once again because launches are expensive and rare, you cannot afford to make mistakes, so everything has to work perfectly
If you are willing to launch to lower orbits, and your launch vehicle is cheap, you are building in bulk, then you can compromise on engineering and accept a few broken sats
Undergrads afaik even high schoolers have built cubesats out of aluminum extrusions, hobbyist solar panels, and a tape measure as an antenna. These things probably dont do that much, but they are up there and they do work.
Pet satellites or cattle satellites?
Kessler satellites
The famous Musk timeline. "By next year, 2 year tops".
By his original timeline we should have landed on Mars… 4 years ago.
https://www.planetary.org/articles/20170929-spacex-updated-c...
and "Full Self-Driving" was a solved problem in 2016.
Ooh, happy 10th anniversary, FSD?
> in what fantasy world
It is already more expensive to performance maintenance on SOCs than it is to replace them. Remember, these machines are not for serving a database, there are practically no storage needs (and storage is the component that fails most often.)
Given that, the main challenge is cooling, I assume that will be figured out before yeeting 100 billion $ of computers into space. Plenty of smart people work at these companies.
Do smart people work at Boring Company? Do smart people work on FSD at Tesla? What about the HyperLoop? It is possible for smart people to make technical achievements without the overall project being particularly successful.
I am super intrigued that Elon thinks this. He's got more information about space based compute deployments than any other person in the world, I think -- every Starlink satellite has a linux box in it -- and to my mind the heat problems and hardening problems seem like they would dominate.
But, apparently not, to Elon. I will say there's a MASSIVE cost to getting power infrastructure, land, legal stuff done on terra firma; all that just sort of .. goes away when you're deploying to space, at least if you're deploying to space early and fast.
At any rate, I'm intrigued. For reference, people like Larry Ellison, another hyper-informed genius business man, are investing in and buying small nuclear reactors for power right now on earth, so we are at some sort of inflection point for datacenter demand. I wonder what the first independent service business is going to be in space. Probably satellite clearing / decommissioning.
Sometimes, I wonder how people in the middle ages accepted the whole "Divine Right" of their ruling Kings, while simultaneously suffering under their rule.
> Larry Ellison, another hyper-informed genius business man
"King George, another royal blessed by the divine."
I don't know if you're aware of this, but American markets are hyper competitive. I'd be extremely wary of any instinct to discount the skill level of any top-20 self-made billionaire industrialist, really anywhere in the world, but in the US at least, that skillset is likely heavily skewed toward business excellence.
> self-made billionaire industrialist
We've reached levels of billionaire worship that would make any court jester of the 1400's blush
An off-the-cuff four word description on an Internet forum definitely exceeds the level of worship from a court jester in the 1400s that had to dress up in costume and dance at the command of a king, lest their head get cut off.
Well jesters are supposed to call out the BS, yes?
That said, How do you (accurately) describe Ellison?
Greedy, ruthless and well-connected? These people are hardly genuises.
lawnmower
> American markets are hyper competitive
Eh. Brand new markets, perhaps. But established markets in the US favor incumbents and encourage monopoly.
Monopoly - so easy everyone can do it. We should give it a go! I would include managing your way into monopoly status and steering clear of being broken up as business skills; no?
The techbro cult is filled to bursting with greedy, narcissistic people who are wholly willing to ignore evil because they expect to be the next dispensers of said evil.
You just responded to one of them.
> I am super intrigued that Elon thinks this.
He has a habit of saying things that ultimately are just hype building. I do not believe that he really believes in space data-clusters.
That's the most measured Elon critique I've read today. :)
I've been told by SpaceX folk that Elon's job is to keep a 20 year view in the future and essentially get folks to work backward from that.
I think I might kind of be sold on data-clusters in space in 20 years.
I can understand if I had lift that cost 1/10 what everyone else in the world paid for it, I'd be even more sold on them.
That said, this newfound enthusiasm of his certainly makes a commercially reasonable path forward to turn xAI stock into spacex stock. Elon takes care of his investors, generally speaking.
I believe that he believe the hype will be good for SpaceX IPO.
Starlink runs special rad-hard computers from AMD. None of that transfers to top of the line GPUs. This is crazy.
SpaceX supposedly mostly runs non-rad-hard parts, the ostensible reason being because its more cost effective to double or triple up than buy specialty equipment. Do you have a source for this?
You are conflating with their rockets. They don’t go space grade in their rockets because they don’t need to. They’re not up there for very long, and the avionics can be shielded behind large rocket stages.
Starlink satellites use space-rated AMD Versal chips: https://www.pcmag.com/news/amd-chips-are-powering-newest-sta...
Google tested the radiation tolerance of tpus which include hbm and they performed fine. https://research.google/blog/exploring-a-space-based-scalabl...
> I will say there's a MASSIVE cost to getting power infrastructure, land, legal stuff done on terra firma; all that just sort of .. goes away when you're deploying to space, at least if you're deploying to space early and fast.
You need both power infrastructure and structures to build within for deploying in space too. And you have to build them and then put it all into space.
Cost per square foot of land is not that high basically anywhere you could build a datacentre to offset that.
Well some stuff you either don't need, or just can't have so you do something different - for instance, transformers to convert grid power - no grid - no transformers. Those are like a 36 month wait list in the US right now. And solar is something like 2x as efficient in space.
I agree those don't seem immediately to be huge wins to me; not dealing with local politics might be a big one, though. Depending on location. There's a lot of red tape in the world.
$/sq foot or meter belies the cost of dealing with every regulatory agency that has claim on that area. There's no environmental commission you've got to pay off if your satellite starts leaking noxious chemicals all over the place, the same way you'd have to if you spilled something at NUMMI in Fremont, California.
We should take a hyperloop trip together, connect our nuralinks and figure this out together. Or perhaps our optimus bots can help us understand?
> He's got more information about space based compute deployments than any other person in the world...
He also had more information about self-driving progress than any other person in the world - yet he was wrong with his predictions every year for last 10 years.
> like Larry Ellison, another hyper-informed genius business man
Don’t anthropomorphize Larry Ellison.
Elon doesn't think this. Elon says whatever bullshit thing comes into his head without regard for technical, economic or physical plausibility. As long as it raises the stock price!
The market has had almost a hundred years of being well-regulated, so when a sociopath lies through their teeth, we're inclined to give them the benefit of the doubt. But in the last few years, that regulation has been worn down to nothing, and the result is and was entirely predictable: fraud.
But spacex isnt a publicly traded company.
This just reads like Elon trying to leverage the AI bubble to prop up SpaceX stock to me.
SpaceX stock needs no propping.
That said xAI might need a bit of a rescue.
It's going to be really funny if this ends up killing SpaceX in the long run.
Launching alone consumes about 75-150kWh per tonne of energy for fuels only (as per ChatGPT).
Planned lifespan of Starlink satellites is 5years.
>This is so obviously false.
One of the biggest but most pointless questions I have about our current moment in history is whether the people in power actually believe the stuff they say or are lying. Ultimately I don't think the answer really matters, their actions are their actions, but there is just so much that is said by people like Musk that strains credulity to the point that it indicates either they're total idiots or they think the rest of us are total idiots and I'm genuinely curious which of those is more true.
We’re at a point where propaganda is so much more powerful than reality that the people in power literally can’t tell the difference. When your source of ethics is the stock price, little details like physical impossibility stop seeming relevant.
Here is my main question: Musk is on record as being concerned about runaway "evil AI." I used to write that off as sci-fi thinking. For one thing, just unplug it.
So, let's accept that Musk's concern of evil runaway AI is a real problem. In that case, is there anything more concerning than a distributed solar powered orbital platform for AI inference?
Elon Musk appears to be his own nemesis.
He just says stuff to convince people of things that benefit him. Internal consistency was never the plan.
My point is not to make fun of him, but to help avoid the destruction of humanity via an HN comment. No joke.
This ish is starting to get really serious.
Aside from anything about Elon Musk, here’s an interesting video response to the “just unplug it” argument on the Compuerphile channel: https://youtu.be/3TYT1QfdfsM
Ha, I figured that might be the video prior to clicking it. I am a long time fan.
Agreed, when I wrote "just unplug it," this counterargument was present in my mind, but nobody likes a wall of text.
However, my original point was that a distributed solar powered orbital inference platform is even worse! Think about how hard it would be to practically take out Starlink... it's really hard.
Now.. >1M nodes of a neural net in the sky? Why would someone who lives as a god, the richest man in the world, the only person capable of doing this thanks to his control of SpaceX... do the literal worst thing possible?
That'd easily take a few LEO detonated fragmentation bombs to trigger a cascading LEO shrapnel field.
It's a lot harder than taking out some terrestrial power lines.
Sure, it'd take obital launch capabilities to lift ... how many bags of metal scrap and explosives?
tone: I don't really understand orbital mechanics, but I do understand geopolitics a bit.
1. China is very concerned about Starlink-like constellations. They want their own, but mostly they want to be able to destroy competitors. That is really hard.
2. Many countries have single ASAT capabilities. Where one projectile can hit one satellite. However, this is basically shoot a bullet, with a bullet, on different trajectories.
3. > Sure, it'd take orbital launch capabilities to lift ... how many bags of metal scrap and explosives?
If I understand orbital mechanics... those clouds of chaff would need to oppose the same orbit, otherwise it is a gentle approach. In the non-aligned orbit, it's another bullet hitting a bullet scenarios as in 2, but with a birdshot shotgun.
My entire point is that constellations in LEO take hundreds of Falcon 9's worth of mass to orbit and delta-v to destroy them, as in-orbit grenades which approach gently. This IS REALLY HARD, as far as mass to orbit, all at once! If you blow up some group of Starlink, that chaff cloud will just keep in orbit on the same axis. It will not keep blowing up other Starlinks.
The gentle grenade approach was possibly tested by the CCP here:
https://news.ycombinator.com/item?id=46820992
> tone: I don't really understand orbital mechanics, but I do understand geopolitics a bit.
Thanks for the clarification, I guess that explains this (from you):
> Think about how hard it would be to practically take out Starlink.
and this:
> My entire point is that constellations in GEO
which you've now corrected.
Moving on:
> My entire point is that constellations in LEO take hundreds of Falcon 9's worth of mass to orbit and delta-v to destroy them, as in-orbit grenades which approach gently. This IS REALLY HARD
So let's not do that .. how hard is it to render the entire LEO zone a shit show with contra wise clouds of frag that cause cascading failures?
Forget the geopolitics of China et al. .. LEO launch capabilities are spreading about the globe, it's not just major world powers that pose a threat here.
Ok... so, let's reset, please. I bet that we have very similar intentions, and yet on internet forums, we have perfected the art of users speaking past each other.
Just to get on the same page here. My arugument is that prior to Elon Musk, the only human capable of launching >1M distributed solar powered inference nodes, if one accepts runaway AGI/ASI as a threat... prior to that we had a few hundred terrestrial AI inference mega-data centers. Most of them had easily disrupted power supplies by one dude with a Sawzall.
Now, we are moving to a paradigm where the power supply is the sun, the orbital plane gives the nodes power 24/7, and the dude with the Sawzall needs to buy >10,000x (not sure of the the multiple here) the Sawzalls, and also give them escape velocity.
Can we not agree that this is a much more difficult problem to "just unplug it," than it was when the potentially troublesome inference was terrestrial?
To address your grade-A snark upthread which provided a chuckle, you may indeed have a better grasp of orbital mechanics than I do; I've not worked in the field professionally since I dealt with raw MODIS streams direct from ground stations.
And, moving on in good faith; yes, things can be difficult but not insurmountable or impossible.
Let's call rendering LEO unusable a real actual risk going orward .. there are papers on this.
My attitudes are shaped by having grown up with literal Z-force / SASR types; my father (still alive) had a lifelong friendship with Major General Philip Michael Jeffery, I took theodolite lessons from Len Beadell, diving lessons from Jack Wong Sue, and my aunts flew spitfires, jets, and heavy bombers in WWII.
There are many people in this world who, if asked, would regard taking out a LEO constellation as an interesting challenge.
What, creating a huge patchwork of self sufficient AIs, forming their own sky based net, seems bad to you, considering the whole torment nexus/Sky Net connotations? It's not like he's planning to attach it to his giant humanoid robot program. Oh. Ohhhhh. Oh no.
Any estimate by Elon musk, you need to add or substract a zero to/from the end. Here, I'll fix it for you.
> The basic math is that launching a 100,000 tons per year of satellites generating 10 kW of compute power per ton would add 1 gigawatt of AI compute capacity annually, with no ongoing operational or maintenance needs. Ultimately, there is a path to launching 0.01 TW/year from Earth. > My estimate is that within 20 to 30 years, the lowest cost way to generate AI compute will be in space.
But more importantly, there is no heat dissipation in space. There’s no atmosphere to cool you, no water you can put heat into. Just an empty void. You can radiate a little, but the sun alone is enough to cook you, without you having a rack of GPUs inside your satellite.
It’s completely delusional to think you could operate a data centre in a void with nowhere to put the heat.
Do you honestly believe that nobody involved has ever considered that?
Apparently not, otherwise this silly idea wouldn't exist.
Naysayers probably get fired fast.
Never mind operational and maintenance costs. In what fantasy world is it cheaper to put a computer in orbit than in a building on the ground? I don't care how reusable and maintenance-free Starship gets, there's no way even absurdly cheap launches are cheaper than a building.
The whole thing makes no sense. What's the advantage of putting AI compute in space? What's even one advantage? There are none. Cooling is harder. Power is harder. Radiation is worse. Maintenance is impossible.
The only reason you'd ever put anything in orbit, aside from rare cases where you need zero-gee, is because you need it to be high up for some reason. Maybe you need it to be above the atmosphere (telescopes), or maybe you need a wide view of the earth (communications satellites), but it's all about the position, and you put up with a lot of downsides for it.
I feel like either I'm taking crazy pills, or all these people talking about AI in space are taking crazy pills. And I don't think it's me.
The most generous interpretation is that the "AI in space" nonsense is a cover for putting limited AI in space for StarShield (military version of StarLink), which is essentially the "Golden Dome".
It might be possible to scam the Pentagon with some talk about AI and killer satellites that take down ICBMs.
You lost me on million tons.
They'll just be decommissioned and burn up in space. nVidia will make space-grade GPUs on a 2-3 year cycle.
They don't need to be space grade, consumer hardware will do just fine.
For AI a random bit flip doesn't matter much.
Only if that bitflip happens somewhere in your actual data, vs. some GPU pipeline register that then locks up the entire system until a power cycle. Or causes a wrong address to be fetched. Or causes other nasty silent errors. Or...
Try doing fault injection on a chip some time. You'll see it's significantly easier to cause a crash / reset / hang than to just flip data bits.
'rad-triggered bit flips don't matter with AI' is a lie spoken by people who have obviously never done any digital design in their life.
As long as they stay below Van Allen belts and deal with weaker magnetic shielding in sun synchronous orbit (high latitudes).
I would say they probably something a little beefier than consumer hardware and just deal with lots of failures and bit flips.
But cooling is a bigger issue probably?
Random bit flips might even improve output.
Single upset events in a modern GPU are not bitflips. They destroy the surrounding circuitry and usually disable the whole unit.
If that happens you disable that CUDA core. If you GPU is too damaged, you deorbit the satellite.
Yeas, and this will happen within weeks of launch with the orbits under consideration.
You do realize that “space-grade” involves process changes that intrinsically incur orders of magnitude efficiency losses? Larger process sizes, worse performing materials. It’s not just a design thing you can throw money at.
> For one thing, in what fantasy world would the ongoing operational and maintenance needs be 0?
Do you not understand how satellites work? They don't send repair people into space.
This has been a solved problem for decades before the AI gold rush assumed they have some new otherworldly knowledge to teach the rest of the world.
> Do you not understand how satellites work?
Not trying to be rude - but it's you who doesn't understand how satellites work.
The U.S. has 31 GPS satellites in orbit right now. The operational cost of running those is $2 million/day.
Not to mention the scale of these satellites would be on the order of 10x-100x the size of the ISS, which we do send people to perform maintenance.
The problem is not solved and the techniques they use to deal with it run directly contrary to maximizing compute, because that's not historically something they have remotely cared about.
I fly satellites. None of them have a zero operational cost. None. Even the most automated cost money to keep running.
The problem is solved by making satellites extremely expensive.
If you had told me 4 years ago that Twitter would be merged into SpaceX I would have called you crazy. Yet here we are..
Is xAI Twitter? I thought they were separate companies, but honestly I don’t know anymore.
Yes X was merged into xAI last year I believe.
Yep.. "all stock deal" last Spring. https://www.paddle.com/news/industry/elon-musk-xai-acquires-...
That’s wild. So it’s true that Twitter is merging with SpaceX?
That's what the headline is, yep. Wild times indeed.
Elon Musk’s AI startup, xAI (x.ai = grok), officially acquired X Corp. (the parent company of the social media platform X or x.com, formerly Twitter/x.com) in an all-stock deal. Both now operate under a unified holding entity, frequently referred to in corporate filings as X.AI Holdings Corp. (or simply xAI). Now SpaceX has moved to acquire or merge with xAI. This effectively brings the social media data from x.com, the AI development of x.ai, and the satellite infrastructure of Starlink/SpaceX under one "super-conglomerate" roof.
I'm just waiting for the year 2350 when Walmart buys Weyland-Yutani.
Thoughts:
1. What in the circular funding? This feels more like a financing scheme founding it under X/Twitter and then spinning it over to SpaceX. I suspect some debt is disappearing or taxes aren't getting assessed because of this move.
2. The only thing harder than harnessing "a millionth of the sun's power" on Earth would be launching enough material into space to do the same thing. And that's not even a reason for SpaceX to own an AI company, at least not at this point. The current AI isn't going to help with the engineering to do that. Right now hiring 20-somethings fresh out of college is way cheaper and SpaceX has been very successful with that.
quick edit: dang, I even got point 1 backwards. xAI owns X/Twitter, and that means that SpaceX now owns X/Twitter as well as an AI company. Super suspicious that SpaceX could actually think that buying the social media part (a significant portion of xAI's value) would be worth it.
Just a neat bit of financial engineering. You can tell because Elon picked SpaceX instead of Tesla – which would have actually made sense at some level (Optimus Robots + AI). But Tesla is public and so he'd need to follow laws and reporting requirements.
You can tell it's just financial engineering because in the entire press release xAI is only mentioned in the first two sentences. Everything after that is Elon talking about space data centers to distract from the actual topic. Which seems to be working
does he need spacex/xai to prop up tesla or the other way around?
Tesla is still very profitable, as is SpaceX I assume. Twitter/X has been a $44 billion dollar failure, and xAI is a vanity project so Musk can go around saying he is a player in the AI space. Investors in both X and xAI need to be bailed out, hence this announcement.
Tesla has a P/E in the hundreds and a ~0.3% market cap to profit ratio. In what world is this "very" profitable?
In the world where it makes $8-10B in profit on $90-95 billion in revenue every year. Whatever price investors choose to trade the stock at is irrelevant to those numbers.
It's actually down to $3.8B in profit now, and will be losing money within a year at the rate its been losing profitability.
2% net return on assets is garbage
The $44B Twitter/X buyout was not a failure. For example Fidelity has its $19M investment in the buyout - now xAI common shares - marked at $62M (up over 3X) as of 12/31/25. It was certainly valued even higher on 1/31/26 after xAI had an oversubcribed fund raise in January. All before this merger announcement.
The fact that it had to be successively bailed out by xAI (which itself was funded by Tesla) and now SpaceX shareholders is exactly what makes the acquisition a failure.
He spent other people's money (or maybe even imaginary money) he couldn't have used for himself (since selling off major stakes in your company is a big nono)
A "bailout" is when a company rescued from bankruptcy. Common equity holders take large losses or are wiped out. This did not happen here.
We also know the Twitter buyout debt was sold at near par before the merger with xAI which is inconsistent with being near bankruptcy.
> xAI had an oversubcribed fund raise in January
My understanding is that it was not oversubscribed and would not have closed without Tesla’s investment.
I think its an effort to position SpaceX as an AI company in order to justify some ridiculous valuation at IPO.
I think it's more so that the upcoming new public shareholders of SpaceX bail out his X/xAI misadventure.
Do you genuinely not think that "Elon" (xAI) is player in the AI space?
You don't have to think they have the best models of course, but they are clearly a very significant, and some might argue, leading player in the AI race.
> and some might argue, leading player in the AI race
What is this argument exactly? What are they leading?
It is a real model, real datacenters, and deployed heavily on their social media platform.
That's the full stack? Only other player that vertically setup is facebook, google and microsoft.
xAI’s models are really not pioneering at all. They weren’t the first to do MoE. Not the first to do open weighting, not the first to have memory or multi-modal vision.
So no, I wouldn’t say Elon is a major player in the AI space. People use his models because they are cheap and are willing to undress people’s photos.
saying they aren't pioneering is very different than saying they aren't a major player in the space. There're only like 5-7 players with a foundational model that they can serve at scale. xAI is one of them
I guess the difference is Tesla is a public company, so requires more paperwork. SpaceX isn't public yet, but will be soon, meaning it will have a cash infusion.
I suspect SpaceX will acquire Tesla at some point. It’s the most profitable of these companies. So basically SpaceX employees and shareholders are covering up for the failing Tesla business and the already-failed xAI business.
Let’s not forget, xAI is the parent of Twitter/X (the social network). So now, taxpayers are paying to keep Twitter/X alive. After all, it is taxpayer money going to the contracts the government gives SpaceX for launches. Nice way to subsidize what is effectively a one sided campaign machine for the GOP and far right.
> I suspect SpaceX will acquire Tesla at some point.
I think that is also likely, unless Tesla can stage a major turnaround, it is going to be beaten by Chinese competitors nearly everywhere that they are allowed (which is everywhere but the USA.)
I get what you're saying, but that taxpayer money is paying for the launch services at a very competitive rate (possibly the cheapest of all available options), not a subsidy scheme.
This was my immediate thought as well. A great time to ask yourself — why am I literally paying for any of this? At best I literally don't use any of these services, at worst they are actively used against me.
Seems like a great way to play games with moving money around. Come up with a "valuation" and then "acquire".
It worked for his Twitter co-investors. I guess they overlap enough with Xai investors for them to think it's more clever than it is a rip off.
I'm pretty amazed one can play a shell game for so long and so openly with the public.
I don't see the demand for space being there, OSS is driving costs down and there are still plenty of hardware and algorithmic optimizations we haven't deployed yet.
> scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!
I didn't realize SpaceX's media press is even cringer than Elon's average tweet...
Microdosing might not be the word for it.
I mean this is clearly directly written by Elon. I suspect the SpaceX comms people are equally eye-rolling.
I am concerned, and haven’t seen anyone else point this out yet, that Musk will move Grok’s CSAM generation capabilities to space to be beyond the reach of terrestrial policing. Does this create some sort of legal loophole here so Musk can do this with impunity?
No. For one, things in space are under the jurisdiction of the country they were launched from. For another, it's people that do crimes, not satellites or LLMs, and the people involved in making CSAM are all on Earth.
AFAIK he can do whatever he wants in space, but CSAM is still illegal to view or even download in most (all?) countries of the world. So unless the degenerates also move out into space (which I'm sure they're eager to do), it wouldn't really ease the legal situation here on Earth.
Ground stations would be the major problem.
Maybe if Elon launched himself and the dev team into orbit and didn't use any ground stations and just Starlink terminals he could start getting into legal loopholes.
So my floating data center in international waters idea has potential investors?
Remember that he wants to make X a bank. An orbital tax haven.
But seriously, I think legally satellites are under the jurisdiction of the country they were launched from.
Well, offshore launches are already a thing.
Or he could just buy a small island in the Carribean. There's one in particular that is available.
That island is part of the US. Technically SpaceX technology cannot be exported to other countries but laws are fake so...
<elon venture> rescues failing <elon venture> here have some <unattainable goals> the shareholders love that shiz.
I still don't understand the "data center in space" narrative. How are they going to solve the cooling issue?
Maintenance cost must be pretty fucking insane
This "Space Datacenter" sounds like biggest bullshit in last decade, which is pretty damn fucking high bar.
Hes committed to building thousands of Optimus robots for a market that does not exist while cutting back on building evs (a market that does).
I think its pretty clear that Musk has lost his goddamn mind. And the American corporate system and Government seem powerless to do anything.
The reason is probably that Tesla is falling behind on EVs, or at least feels like they've juiced all they can from them at the moment, but advanced robotics is still on the upswing and probably is far from reaching its full potential. They have enough money that moonshots like these probably seem irrelevant at their scale.
As for the space datacenter idea, I think this is just a case of extreme marketing that Musk's ventures are so accustomed to. Making huge promises to pump their stocks while the US government looks the other way. When time comes for them to deliver on their promises, they've already invented ten more outrageous ideas to make you forget about what they promised earlier. Hyperloop as a viable mode of transportation, tunnel networks for Teslas, SpaceX vehicles as a mode of transport, X as the new 'everything app', insane timelines for a Musk-led human mission to Mars. They've done it all.
To be precise: humanoid robot TAM $0; vehicles TAM ~$2.7 trillion.
There is no maintenance, you have many cheap satellites - if one fails you just deorbit it.
How many is "many"?
One million.
You think I'm joking but I'm not. https://spacenews.com/spacex-files-plans-for-million-satelli...
I think it's fair to say past 30 years. Dotcom boom only had modest cons by contrast
Send up a spacecraft with back-to-back / equal area solar panels and radiators (have to reject heat backwards, can't reject it to your neighboring sphere elements!). Push your chip temp as much as possible (90C? 100C?). Find a favorable choice of vapor for a heat pump / Organic Rankine Cycle (possibly dual-loop) to boost the temp to 150C for the radiator. Cool the chip with vapor 20C below its running temp. 20-40% of the solar power goes to run the pumps, leaving 60-80% for the workload (a resistor with extra steps).
There are a lot of degrees of freedom to optimize something like this.
Spacecraft radiator system using a heat pump - https://patents.google.com/patent/US6883588B1/en
Same way the dude solved Roadster, full self driving etc.
Presumably the cooling problem gets hand waved away as a technical detail, and the real selling point is data centers that aren't subject to any regional governments laws.
> and the real selling point is data centers that aren't subject to any regional governments laws.
No? ISS isn't exempt from legal systems.
https://www.esa.int/Science_Exploration/Human_and_Robotic_Ex...
So the csam generator can live on? Why not just pirate radio style it and cool it with ocean water....
Periodically send a crew to a comet to bring back a large slab of ice to put in the data center.
Cooling a datacenter in space isn't really any harder than cooling a starlink in space, the ratio of solar panels to radiating area will have to be about the same. There is nothing uniquely heat-producing about GPUs, ultimately almost all energy collected by a satellite's solar panels ends up as heat in the satellite.
IMO the big problem is the lack of maintainability.
> Cooling a datacenter in space isn't really any harder than cooling a starlink in space
A watt is a watt and cooling isn't any different just because some heat came from a GPU. But a GPU cluster will consume order of magnitudes more electricity, and will require a proportionally larger surface area to radiate heat compared to a starlink satellite.
Best estimate I can find is that a single starlink satellite uses ~5KW of power and has a radiator of a few square meters.
Power usage for 1000 B200's would be in the ballpark of 1000kW. That's around 1000 square meters of radiators.
Then the heat needs to be dispersed evenly across the radiators, which means a lot of heat pipes.
Cooling GPU's in space will be anything but easy and almost certainly won't be cost competitive with ground-based data centers.
Sure, but cooling a starlink in space is a lot more difficult than cooling a starlink on earth would be. And unlike starlink which absolutely must be in space in order to function, data centers work just fine on the ground.
This. There's no scenario where it's cheaper to put them in space.
I think there's a lot of a room for an energy play that will ultimately obviate the enormously costly terrestrial energy supply chain.
You can just use the cheap solar panels that were gonna be launched into space (expensive) and not launch them into space (not expensive) and plug them into some batteries (still, cheaper than a rocket launch)
According to Gemini, Earth datacenters cost $7m per MW at the low end (without compute) and solar panel power plants cost $0.5-1.5m per MW, giving $7.5-8.5m per MW overall.
Starlink V2 mini satellites are around 10kW and costs $1-1.5m to launch, for a cost of $100-150m per MW.
So if Gemini is right it seems a datacenter made of Starlinks costs 10-20x more and has a limited lifetime, i.e. it seems unprofitable right now.
In general it seems unlikely to be profitable until there is no more space for solar panels on Earth.
His bet then, is that the $1 million cost to get a Starlink V2 mini into orbit can be made cheaper by an order of magnitude or two.
I think that it's not just about the ratio. To me the difference is that Starlink sattelites are fixed-scope, miniature satellites that perform a limited range of tasks. When you talk about GPUs, though, your goal is maximizing the amount of compute you send up. Which means you need to push as many of these GPUs up there as possible, to the extent where you'd need huge megastructures with solar panels and radiators that would probably start pushing the limits of what in-space construction can do. Sure, the ratio would be the same, but what about the scale?
And you also need it to make sense not just from a maintenance standpoint, but from a financial one. In what world would launching what's equivalent to huge facilities that work perfectly fine on the ground make sense? What's the point? If we had a space elevator and nearly free space deployment, then yeah maybe, but how does this plan square with our current reality?
Oh, and don't forget about getting some good shielding for all those precise, cutting-edge processors.
Assuming you can stay out of the way of other satellites I'd guess you think about density in a different way to building on Earth. From a brief look at the ISS thermal system it would seem the biggest challenge would be getting enough coolant and pumping equipment in orbit for a significant wattage of compute.
Why would you need to fit the GPUs all in one structure?
You can have a swarm of small, disposable satellites with laser links between them.
Because the latencies required for modern AI training are extremely restrictive. A light-nanosecond is famously a foot, and the critical distances have to be kept in that range.
And a single cluster today would already require more solar & cooling capacity than all starlink satellites combined.
Because that brings in the whole distributed computing mess. No matter how instantaneous the actual link is, you still have to deal with the problems of which satellites can see one another, how many simultaneous links can exist per satellite, the max throughput, the need for better error correction and all sorts of other things that will drastically slow the system down in the best case. Unlike something like Starlink, with GPUs you have to be ready that everyone may need to talk to everyone else at the same time while maintaining insane throughput. If you want to send GPUs up one by one, get ready to also equip each satellite with a fixed mass of everything required to transmit and receive so much data, redundant structural/power/compute mass, individual shielding and much more. All the wasted mass you have to launch with individual satellites makes the already nonsensical pricing even worse. It just makes no sense when you can build a warehouse on the ground, fill it with shoulder-to-shoulder servers that communicate in a simple, sane and well-known way and can be repaired on the spot. What's the point?
Starlink already solved those problems, they do 200 GBit/s via laser between satellites.
And for data centers, the satellite wouldn't be as far apart as starlight satellites, they would be quite close instead.
that's the best part! They don't!
Remember MoviePass, and how they were losing gobs of money by letting people see unlimited movies for $20/month?
It was so obviously stupid that a bunch of people went, "well, this so clearly can't work that they must have a secret plan to make money, we'll invest on that promise", and then it turned out there was no secret plan, it was as stupid as it looked and it went bankrupt.
The "datacenters in space" thing is a similar play: it's so obviously dumb that a bunch of smart people have tricked themselves into thinking "wow, SpaceX must have actually figured a way it can work!"; SpaceX has not and it is in fact exactly as stupid as it looks.
But it won't end the same as MoviePass until Elon dies; he will keep moving things around, propping up failures with VC, IPO, federal/state (taxpayer) and profit making business money.
My guess would be just a regular radiator and cooling system like a liquid pump. The only obstacle should be the vacuum. That said I don't have any hopes Elon has any understanding of any of it.
Too me this smells of projected cash desperation. Do people actually pay for Grok?
You get Grok with paid x.com ; so there is some sort of cash from that, I would guess.
Is Grok driving any x.com subscriptions?
There are regularly comments on HN from people who say they pay for Musk's various products, and I am always downvoted into oblivion for suggesting that that is ethically problematic.
There's obviously quite a lot of autocratist illiberals in tech.
as you can see from the Epstein files, people with minor interests have a lot of money
> In the long term, space-based AI is obviously the only way to scale.
I never questioned it.
Space is also extremely cold, and if it's as dense as Musk cooling won't be an issue.
I can't tell how many layers of sarcasm are here, but I just want to highlight that aktshually cooling in space is quite difficult because there is no convection, so the only cooling option is radiative. Which gets a bit hard when the satellite gets blasted by the sun.
The ISS doesn't have problems staying warm, it has problems cooling off.
> the only cooling option is radiative.
It does say he's planning an AI sun, I'm guessing that's the temperature you need to run at for radiation to work.
> It does say he's planning an AI sun
Everything I've heard from Musk in the past decade has been against my will and has made me dumber. (no I do not care to verify or know whether the above is true)
Edit: ah fuck ya got me "the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe" what the cultish bullshit is this. In a just world investors would be fleeing in droves from this cuckoo behavior (I know xAI & SpaceX are private)
There are already large communication satellites that consume several kW of power.
Oh, good. So we only need to multiply that by 200 million times, per space datacenter.
The data center would still consist of many individual satellites, much like a earth based data center consists of many individual servers
A large telecommunications satellite operates at about 15kW. A Blackwell GPU consumes 1kW so you would be at 15 Blackwells per satellite. The cooling surface needs to scale linearly so there is little return to scale.
This doesn't sound like a good idea to me.
tbh you could just combine them with starlink sats. didn't they just apply for (and get?) a license for 1 million sats? Stick a single racks worth of gpu power on those and hey presto you've just got yourself the largest ai cluster in the world by far.
Hence the “dense as Musk” comment
isn't this just fraud in broad daylight? I don't get it. Why not at least try to hide it?
It's crazy but legally there's nothing fraudulent here. I'm sure the deal was approved by the boards of both companies.
Its only fraud if poor people do it. Welcome to American politics.
But the "fraud" here is being done mostly to VC investors with deep pockets and lawyers, at least until he tries to take this entity public. And I can't imagine them just taking this lying down, but then again maybe they realize that offloading this steaming pile on public market investors is the best way out. But even then... SpaceX seemed like it was quite viable on its own, the investors there are the real losers here.
It is all very puzzling to me.
no I get it, but I mean fraud is usually kept out of the public. this is fraud in broad daylight?
Why hide it if you know you won't be punished for it?
Poor people are using their public car company to buy their private space company?
5 days after Tesla gave xAi 2 billion.
Funnel money from a public company to a private entity and then make it disappear. Poof! Magic.
> Starship will deliver millions of tons to orbit and beyond per year
Excuse my naive physics, but is there a point at which if you take enough mass off of earth and launch it into space, it would have a measurable effect on earth's orbit? (Or if the mass is still tethered to earth via gravity, is there no net effect?)
Earth weighs about 5972200000000000000000000 kg. They are claiming to plan to launch ~5000000000 kg / year. That's 8x10^-14 %. You'd need some pretty accurate instruments to tell the difference.
Is there anything substantially different about Google's announcement https://news.ycombinator.com/item?id=45813267 that makes it any more sane than the Space-X announcement?
There is literally an emoji in the middle of the announcement post. Very on brand for Elon.
Related: NASA chief suggests SpaceX may be booted from moon mission [1]. Blue Origin could snatch SpaceX's Starship lander contract. This looks increasingly a good idea.
[1] https://www.cnn.com/2025/10/20/science/nasa-spacex-moon-land...
Sean Duffy is no longer acting administrator of NASA. This proposal was apparently part of a bid to get the support of a coalition of old-space companies and new-space non-SpaceX companies. As part of that strategy he apparently leaked Isaacman's Project Athena document and was backgrounding that he was a SpaceX plant.
But, Isaacman is administrator now, and whatever you think about Isaacman and his relationship to SpaceX, I don't think there's much merit in thinking one of Duffy's half thought out plans is likely to be carried out.
Sadly this seems correct. When Trump was re-elected Elon Musk pushed for Jared Isaacman to be appointed as NASA administrator. When the pick went another way, it led to some real friction between Musk and Trump. Now, with Isaacman finally at the helm of NASA, it looks like Musk’s influence over the agency has come full circle.
> This looks increasingly a good idea.
Why?
The reason for the entire moon mission is national prestige.
Is financial fraud consistent with our national prestige?
There are better companies.
The original Moon mission was masterminded by a literal card-carrying ex-member of the Nazi party (Wernher von Braun) and the American public back then didn't seem to mind.
SpaceX hits delays according to the article
And blue origin has no delays because they don't even launch anything.
https://www.reuters.com/science/blue-origin-launches-new-gle...
The really skeptical take here is that eventually all of Musk's companies merge, or at least the biggest ones, for juicing that market value to get that $1T payout. Looking at Tesla.
Then he will spin off and then remerge again. He has several more small companies before he plays that game. Maybe next quarter!
Seems like a way to put a lot of junk in space. If thats in earth orbit it will lead to a lot of junk falling from the sky in 10 years. If it all burns up that will be a lot of nasty shit in the atmosphere - millions of tons!
So... Elon wants to literally build Skynet?
This is definitely better than merging with Tesla.
They can sell xAI/Grok to all automobile companies along with Tesla and other businesses(X.com included) just like the SpaceX services.
It would good to see how it was valued.
https://www.spacex.com/updates#xai-joins-spacex additionally the longer article on SpaceX site
Kind of a bad look - but I can't precisely say why. Maybe he thinks he can raise more capital this way than he could for each company separately? Especially raising more money for X might be quite hard - they seem to be quite a bit behind on the revenue side compared to OpenAI / Anthropic. With both companies merged he might just find enough retail investors willing to buy at sky high valuations.
Doesn't the idea of Orbital Datacenters imply that the constraining resource right now is physical space, and not compute, electricity, etc?
Did we suddenly solve the electricity problem, or the compute problem? As far as im aware there are still plenty of datacenters being planned and built right now.
"Thou shalt not make a machine in the likeness of a human mind."
And so it began. The seed was sent into space. All going according to plan.
Accountants will be studying the deals and cyclical valuations of AI companies in the same way we study bank runs and FDIC insurance today.
> buy a dying social network for 44bil
> merge it with a company created out of thin air for 20bil.
> have a third company buy it.
put it back on the market for 1.5 trillion.
[flagged]
One thing to keep in mind. xAI and SpaceX both have contracts with the DoD. So it makes sense he moved it there rather than Tesla. Not sure I buy the needing AI for doing more in space or if this is to save sinking ship, but if one of his two big companies needed to buy it to keep it afloat it makes sense it was SpaceX and not Tesla.
I'm wondering if SpaceX's going public will be delayed. If not we'll see the first test of the public's appetite for what the AI companies' balance sheets look like
Genuine question: is it even theoretically possible to find some way to dump the heat that would be generated by a "data center" in space?
Yes, satellites and the ISS successfully radiate heat today.
Whenever computer chips go into space, they have to be hardened against radiation, because there is no atmosphere to protect them. Otherwise you get random bit flips.
This process takes a while, which is partly why all the computers in space seem out of date. Because they are.
No one is going to want to use chips that are a many years out of date or subject to random bit flips.
(Although now it got me thinking, do random bit flips matter when training a trillion parameter model?)
not a problem for "AI". it's just a bit more spice (temperature) which Grok possibly need! jk!
I know that per HN's guidelines we're supposed to be "kind and curious", and "reply to the argument instead of calling names". But with some texts, engaging with individual arguments loses sight of the more important bigger picture. So while unkind, the most "thoughtful and substantive" thing I think can be said about this text is:
The man's a moron.
They must have linked the wrong press release /s. I would have expected a press release about SpaceX acquiring xAI to talk about why they did that. Or at least mention xAI beyond the first paragraph. This is just Elon talking about space data centers
Didn’t Elon say that orbital solar collection was a stupid idea due to energy loss in transmission? Using AI as an almost proof-of-work shows that it may potentially be more complex problem than previously thought. If we threw Bitcoin miners up to those satellites you could literally beam money down.
Musk is moving value out of public hands and into his own. He overpaid $44B for Twitter, then rebranded it as an AI asset by folding X into xAI. He pushed Tesla to invest $2B of shareholder money into xAI despite shareholders voting no. Five days later, SpaceX acquired xAI, effectively turning Tesla’s cash into equity in a private company Musk owns far more of. Musk controlled every step, there was no real arm’s-length process, and he almost certainly knew the outcome in advance. Musk and his private investors get control, inflated valuations, and IPO upside. Tesla shareholders supply the cash, take the risk, and lose leverage.
Just checking (genuine question) there wouldn't be a sneaky way to weaponize a million satellites in orbit around the Earth, would there? I can't imagine it wouldn't have ever been looked into.
This makes me genuinely sad. SpaceX was the one thing of his that Elon has largely avoided screwing up. Imho, this is in large part due to Gwynne Shotwell. She seems to have the personality (not to mention, personal wealth) to kick Elon in the head when he tries to mess things up.
What’s happening now is nothing more than a transparent effort to couple the AI hype-wagon to SpaceX in order to drive the valuation higher in the minds of investors who still think that LLMs will completely transform society.
I’ll be thrilled if the rocket folks can avoid being distracted by this nonsense, but I’m not optimistic.
I’ve been following SpaceX since something like the 2nd Falcon 1 launch and this is the worst thing I’ve seen happen. Sad times.
I think it’s just financial, I don’t see this as being detrimental or disruptive to SpaceX much at all.
> My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space.
I have never been so tempted to join Kalshi
Let's call it for what it is, a payday for Elon. Paper billionaires have figured out they cannot cash out with out tanking their paper, so now you have these circular deals to extract as much as possible. If we had a functioning government they would step in and put an immediate stop to this on national security grounds.
Elon investors should try buying a lottery ticket, it also lets you dream of the future while not providing returns.
I suspended my disbelief and gave it a chance but I couldn’t hold it anymore after the emoji.
I've yet to attain full-stack mastery in my job, but Musk has already attained capital stack mastery.
I suppose one of the ADR’s read something like “…who cares about bitflips, man. Isn’t AI all about probability?”
Knowing the insane level of hardening that goes into putting microcontrollers into space, how to the expect to use some 3nm process chip to stand a chance?
A trend at the moment is to just hope for the best in cubesats and other small satellites in LEO. If you’re below the radiation belt it’s apparently tenable. I worked somewhere designing satellite hardware for LEO and we simply opted to use consumer ARM hardware with a special OS with core level redundancy / consensus to manage bit flips. Obviously some problems will present for AI there… but there are arguably bigger problems with AI data centres like the fact that they offer almost no benefit with respects to the costs of putting and maintaining stuff in space!
I have so many conflicting thoughts that I cannot properly articulate yet. I can say though, this is not going to end well for most, it is clumsily premeditated and starting to feel like dude is just trying to be a Neal Stephenson character.
This is either insanely ambitious genius or pure shithousery. I guess we'll find out which one it is in 10 years
Given some of Musk’s previous statements, I think I know which one it is already.
well, Musk has been overpromising and under delivering for a decade (or more?), so it seems pretty clear this too is shithousery, albeit possibly ambitious.
Is he also talking about moving X's servers (since xAI owns X) into space?
I will not be left holding this bag. This is such financial engineering nonsense, and if we had any sort of regulatory controls this would never be allowed to happen - especially BECAUSE of national security reasons.
Any details regarding valuations etc?
This makes a lot of sense. The commercial launch business is not large enough to support all possible Falcon launches, so Starlink was created to take advantage of the low launch cost and vertical integration and is now a major profit center for SpaceX.
Starship launches are only going to make sense every 779.94 days (the approx 2 year Mars-Earth proximity). The rest of the time, the launches could similarly be used to deploy orbiting data centers for XAi/Grok etc. Brilliant move.
Do this math include the cost and weight of the radiators? Because it obviously can't work without big radiators, and I don't see them mentioned in the math?
Perfect timing to offload that debt onto the bag— I mean shareholders.
What if enemy or some anti-ai activities or etc attack it? How to protect it? It's just a too easy target.
It's just a dumbest idea ever if Elon truly believes it. I'm pretty sure he doesn't.
I hope all the Tesla shareholders understand that they’re about to get hosed.
Musks making Tesla seem like a good fit into the portfolio.
Friendly reminder for anyone that forgot - xAI acquired Twitter, so now Space-X is the proud owner of a dying social media platform that they overpaid for.
Any claims that this is about putting compute in space is just a non-sense distraction. This was absolutely about bailing Elon out of his impulsive, drug-fueled Twitter purchase.
The only question now is: when they try to go public, will they be punished for wasting so much money or not? My guess is: not.
Has SpaceX figured something out related to photonic chips that dramatically reduces waste heat generation of compute?
Makes about as much sense as Twitch buying Curse about.. a decade ago?
SpaceX has jumped the shark.
The first M&A announcement I've seen in my entire life that includes a laughing emoji; maybe that's what it is!
Reminder that SpaceX has received an estimated $38 billion in government funding over the years, and all of its returns are going to a small set of private investors.
Socialized losses, privatized profits. As is the American way.
Besides the obvious, why do engineers with real skills still work for this guy?
We just had an X8.1 CME event, I just want to point out that at any moment we could have an x40 (we had a carrington event already in 1859) or higher event and all those sats at low earth orbit would be fried and start hitting each other, if SpaceX keeps launching more it becomes incredible probable that we might hit the Kessler Syndrome, and we would legit lose access to Space for a WHILE, including all of what satellites entail.
Are we ready for that as a modern society or are we going to start enacting regulation against it? I'm sorry but people wanting internet everywhere does not justify we going back to the dark ages for a decade or more.
US is dialing back even on global warming. There is no chance current government would risk space superiority for some kessler-shmessler that nobody has ever seen.
If Musk and SpaceX are serious about putting 1 million datacenter satellites into space, then they are not serious about Mars.
You cannot simultaneously build and launch 10’s of thousands of Starships to deliver 1 million tons of equipment and supplies bound to Mars while also committing to launching 10’s of thousands of Starships to orbit full of satellites.
They would need to quadruple their launch rate, and half of those launches would be Starships bound for Mars, the vast majority of which would never return.
How many Falcon9’s have ever been built? It is incredible to say you can build that many rockets and use up that much fuel on any reasonable time scale. You might as well say the Tesla Roadster version 3 will be a Single Stage to Orbit rocket car.
Maybe this is beyond parody…
https://x.com/elonmusk/status/1005577738332172289
Self-driving single stage to orbit rocket cars. It can drive you to work and then go on a Starlink launch to LEO and be back in time to pick you up.
Add your car to the SpaceX fleet and get paid to own a Tesla!
Why would you put data centres in space?
Seemingly to prop up the value of your companies
Right, just to meet the most minimal of the scenarios for datacenters, someone upthread has calculated "launches every 9 hours, 24/7" as a minimum.
Why?
Because it's another tool to move money on books and make it seem that spaceX and or xAi look good to investors when needed. That would be my guess.
Pre-IPO price padding. xAI is going nowhere but at least for now it has some value. Move it under SpaceX, bump up SpaceX’s valuation and therefore it’s opening IPO price. Then kill xAI and write it off.
These are no longer tech companies, to they are financial, and power, instruments of the billionaire class
> These are no longer tech companies, to they are financial, and power, instruments of the billionaire class
SpaceX has made numerous breakthroughs in reusable launch vehicles, human spaceflight, satellite constellation, and rocket propulsion.
SpaceX is the world's dominant space launch provider with its launch cadence eclipsing all others, including private and national programs.
All as a side effect...
Huge drag we allow it to be controlled by a drugged out nut job.
Purely financial shenanigans. Nothing to see here, please move along.
Disgusting.
So Elon has more shares when SpaceX IPOs?
> Current advances in AI are dependent on large terrestrial data centers, which require immense amounts of power and cooling.
You know what's even harder to cool?
> Orbital Data Centers
I really would like to see a cost and cooling breakdown. I just can't see how you can do radiative cooling on the scales required, not to mention hardening.
I thought this was a troll by Elon, now I'm leaning towards not. I don't see how whatever you build being dramatically faster and cheaper to do on land, even 100% grid independent with solar and battery. Even if the launch cost was just fuel, everything else that goes into putting data centers in space dwarfs the cost of 4x solar plus battery.
Nobody knows cooling satellites better than SpaceX
I cannot really tell satire apart from genuine opinions anymore.
(But I do hope it was satire, if not, cooling satelites was/is a big issue and they only have very modest heat creation. A data center would be in a quite different ballpark)
Maybe so, but the actual SpaceX engineers are powerless to stop Elon running his mouth.
This is basic physics lol
Perhaps SpaceX incentive is to lie?
> the world’s foremost real-time information and free speech platform
What a joke.
Now he just needs to work in crypto satellites down to users via all the new phones supporting satellite link to SpaceX. I kinda expected that one first. Distributed payment network outside of government control/oversight seems like something he would be in to.
What kind of financial engineering is going on here? Is xAI about to go bankrupt or something?
Probably. They are building the largest data centers in the world with very little revenue.
So they use a valid and valuable company to hide a giant dumpster fire company. To add to that, their best argument is "AI in space", which has some real "solar roadways" energy to it. I honestly don't know how any SpaceX shareholder could approve this.
I asked Gemini for a two word summary and it wrote "financial engineering"
Grok: Integrated Ambition
The next step will be merging SpaceX and Tesla.
Tesla has probably the most valuable shareholders on Earth. Over years of empty promises and meme status, the stock has pretty much purged all the level heads. So it's mostly deluded Elon sycophants giving placing their tithe on the alter of his sci-fi fantasy smoke and mirrors game.
In reality he will be dumping the debt of twitter and xAI (and maybe spacex?) on Tesla shareholders, and buoying that with the added layer of hyper that spaceX brings.
Musk earns a $1tn payout when Tesla hits $8.5tn dollars.
I expect the next step in this series of moves is to turn Tesla into a SPAC & have it acquire SpaceX, bringing its valuation nearer that 8.5t.
Pretty terrible for SpaceX. Of course they paid a crazy inflated price for xAI in an attempt to cash in on the IPO. This just devalues SpaceX and exposes the investors to all the AI bubble risk.
> SpaceX has acquired xAI to form the most ambitious, vertically-integrated innovation engine on (and off) Earth, with AI, rockets, space-based internet, direct-to-mobile device communications and the world’s foremost real-time information and free speech platform. This marks not just the next chapter, but the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!
I think Elon's taken one too many puffs of hopium
You'd think he'd have a pretty huge tolerance at this point.
So he doesn't want to go to mars he wants to make a big space chatbot?
Plus I suspect the sun is already sentient no need to reinvent the sun
> I mean, space is called “space” for a reason. [Face with Tears of Joy]
Reminder that space only allows for radiative cooling (since there is no air to absorb heat) so data centers in space are going to have massive cooling panels.
xAI owns Twitter... So now space company owns Twitter? Wtf
What this tells me - xAI is essentially a failure, though at what level I'm not sure.
What about security?
The way I read it, X AI is not really profitable and Elon's creditors/coinvestors asked for something tangible for their money, a.k.a shares in SpaceX, his only business that still has some solid foundation. The rest is emois.
Terrible news for SpaceX.
Major grift vibes, inventing half-baked reasoning to justify massive valuations. If money wasnt so deeply entwined with politics at this point, this is the sort of news that would launch fraud investigations.
Discussion on previous speculation: https://news.ycombinator.com/item?id=46814701
It's a CONSTANT stream of new ideas with no payoff at this point.
Hyperloop > Neuralink > Self-Driving Cars > Robotaxi fleets > Personal Robots > Orbital Datacenters > [insert next vibe shift]
At what point do people start to see the ever-shifting goalposts for what they are?
Hyperloop was never a company project, neuralink was a separate company, tesla is rolling out driverless robotaxis and fsd is amazing, data centers seem, robots inevitably are going to do the majority of work - there’s no real doubt about it is there?
Datacenters in orbit seem insane so idk we’ll see
> robots inevitably are going to do the majority of work - there’s no real doubt about it is there?
There's a lot of doubt that the AI and compute to enable that would happen on commercially relevant timescales.
Consider: "do the majority of work" is a strict superset of "get into car and drive it". The power envelope available for an android is much smaller than a car, and the recently observed rate of improvements for compute hardware efficiency says this will take 16-18 years to bridge that gap; that plus algorithmic efficiency improvements still requires a decade between "car that can drive itself" and "android that can drive a car". (For any given standard of driving).
And that's a decade gap even if it only had to do drive a car and no other labour.
You can't get around this (for an economy-wide significant number of androids) by moving the compute to a box plugged into the mains, for the same reason everyone's current getting upset about the effect of data centres on their electricity bills.
And note that I'm talking about a gap between them, not a time from today. Tesla's car-driving AI still has safety drivers/the owner in the driving seat, it is not a level 4 system. For all that there are anecdotes about certain routes and locations where it works well, there's a lot of others where it fails.
That said: Remote control units without much AI are still economically useful, e.g. a factory in Texas is staffed entirely by robots operated over a Starlink connection by a much cheaper team in Nairobi.
Don’t forget LEO satellite internet! Oh wait….
What does Gwynne Shotwell think of this. She seemed level-headed, but is she also batshit insane now by osmosis?
Musk always merges his companies when one is suffering:
Twitter/X in xAI
SolarCity into Tesla
xAI into SpaceC
I am just waiting now for Tesla to be acquired by SpaceC as it has run into issues.
knowing elon, he will make this actually work, thus fully vindicating both the financial engineering and his arrogance!
Elon Musk is a genius, but he’s a financing genius. Look at the long history he has of false promises supporting financing deals between his companies and you’ll see this for what it is, a cash injection and a lie to justify it. He did the same thing with a fake solar roof demo when Tesla bought the almost bankrupt Solar City. He also shifted resources from Tesla and SpaceX to support X in the early days. Even founding xAI outside of Tesla, when so much of its valuation was built on its AI capabilities, was questionable.
I thought this wasn't viable due to cooling requirements - how do you cool massive amounts of compute when the only option is to radiate it into space - nothing to convect it with?
Also, the incredible amount of grift here with the left hand paying the right is scarcely believable. Same story as Tesla buying Solarcity. Board of directors should be ashamed IMO.
Yes. It is very cold up there but there is also no matter, or very little matter. So head conduction and convection don't work, it's all radiation. When we are learning to solve heat transfer problems in engineering school we are generally taught to neglect radiation, because it's effect on cooling the system is typically second or third order when compared to the to "big C's"
It would take roughly 5000 square meter area to cool a typical small data center heat output (1 MW). Not great, not terrible.
Apparently, OpenAI plan to build 250 GW of computing capacity by 2033.
To put that in space, based on your numbers, that's 1,250 square kilometers of cooling - an area roughly equivalent in size to Los Angeles
That is a very tiny amount of compute though.
Cooling and maintenance (part swaps, etc.) are one of many obvious reasons why this is bullshit.
Doesn't stop grifters, tough.
in actual datacenters you often don't even bother swapping parts and just let things die in place until you replace whole racks
Not my experience at a hyperscaler, at least a while back. It definitely made financial sense to swap a small part to get a ~50-100k$ server's capacity back online.
When does the market realize this is all just a shell game and the emperor really has no clothes?
We saw this on a much smalelr scale a decade ago when one of Elon's companies (Tesla) acquired a second one of Elon's companies (SolarCity) because it was broke and owed a ton of money to a third one of Elon's companies (SpaceX).
Elon was forced to go through with his impulsive Twitter acquisition by a Delaware court, an acquisition that was not only secured by a bunch of Tesla stock but also a bunch of Qatari and Saudi royal money. He then mismanaged Twitter so badly Fidelity wrote down its value by at least 80% [1].
So what did Elon do? Raised even more questionable foreign money into xAI, diverted GPUs intended for another of his companies (Tesla) into Twitter and then "merged" Twitter into xAI, effectively using other people's money to bail him out from an inevitable margin call on his Tesla stock.
Interestingly, Twitter was reportedly valued at $33 billion in this deal [2], significantly more than the less than $10 billion Fidelity valued Twitter at. Weird, huh? With a competent government, this would be securities fraud that would have you spend the rest of your life in jail. And even with all that, $11 billion was lost on the deal.
So here we are and it's time for the shell game to be played again. Now it's SpaceX's turn to bail out the xAI investors.
And what is the argument for all this? AI data centers in space. Words cannot describe how little sense this makes. Launch costs (even if the Starship launch costs get to their rosy projections), cooling in space, cosmic rays (and the resulting errors) and maintenance. Servers constantly need parts replaced. You can just deorbit the satellite instead but that seems like an expensive way of dealing with a bad SSD or RAM chip.
[1]: https://www.cnn.com/2024/10/02/business/elon-musk-twitter-x-...
[2]: https://www.cnbc.com/2025/03/28/elon-musk-says-xai-has-acqui...
fool me once (Solar city) shame on you, fool me twice...
hmmmm
> This marks not just the next chapter, but the next book in SpaceX and xAI's mission: scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!
One of the dumbest things I've ever read.
Making a "sentient sun" is the most bald-faced asinine, drug-induced nonsense that should be the complete destruction of all credibility to anyone who said it or typed it or works for anything here.
Who got the money? Hahah
The whole universe was supposed to be turned into paperclips, now it is being turned into graphics cards to produce images of barely legal girls on X.
And Musk keeps grifting about Kardashev 2 civilizations while his rockets do not even reach the moon.
If SpaceX goes public, that will rescue his xAi shares. I wonder how he will rescue his Tesla shares.
What metric does reaching the moon fulfill?
Nothing, save for advertising that you can. And Musk obviously can't, or he would have by now.
Financial engineering. Twitter under Elon became a dumpster fire of porn and hate and big banks were holding 13B in bonds that wouldn’t be worth the paper they were printed on for the company alone so he just links it with his only company that actually is doing something worthwhile…
Not sure how X which “merged” wit X (formerly Twitter) and SpaceX really matter or synergize but here we are. It’s all about the money being protected. And this Ketamine using wierdo is gonna be the worlds first trillionaire. Yay all of us.
> By directly harnessing near-constant solar power with little operating or maintenance costs, these satellites will transform our ability to scale compute. It’s always sunny in space! Launching a constellation of a million satellites that operate as orbital data centers is a first step towards becoming a Kardashev II-level civilization, one that can harness the Sun’s full power, while supporting AI-driven applications for billions of people today and ensuring humanity’s multi-planetary future.
Apparently optimus robots don't work and he needs to start his final grift, space datacenters, while his datacenters on earth are powered by gas turbines.
Most likely he's just trying to bury his epstein involvement where was exposed lying by his own daughter.
Musk: "What do we have that OpenAi doesn't have?"
Musk aide while high: "sPaCe"
Slightly less high Musk aide: "But what is the synergy, where's the moat and how could that be done in practice and most importantly is there any limiting factor on Earth before we have to bring AI into.."
Musk : "SPACE!!!!!!"
It is incredible to think that the extremes of the stock market are actually pretty similar, pink sheets/cryptos and these mega companies are actually the same. News fueled pumps and dumps to win the cycle of hype of the week
They were all inside his head or bots replying to him on Xitter
Ahahaha, who got the money?
Going to be marked at some delusional valuation and at IPO retail bag holders are going to get absolutely massacred.
Like Musk's other great public company failures?
They are failures as far as products
Tesla is marked at purely delusional prices as well. Total hopium on Optimus taking off while their core business craters.
wow this is one of the most incredibly fraudulent things that ever happened in American capitalism and I'm not ignoring Enron or the mortgage meltdown. I'm speechless the US has given up on any semblance of law and order in matter of financial markets and this stuff can happen without people going to jail.
BUut bUt bUt iSn'T TwItTeR / X dYiNg? wHy hAsN'T iT cOmPlEtEly cOllApsEd yET?
The X and xAI doomsters are in complete shambles in an absolute failed prediction of the collapse of the November 2022 acquisition of Twitter.
Instead it is now part of SpaceX, still running with well over 240M+ daily actives and 500M+ MAUs.
https://x.ai/news/xai-joins-spacex wrong link?
Well surely this acquisition is above board. Nothing funny going on here, just good old business as usual.
> Well surely this acquisition is above board.
What makes you think it isn’t?
There's an epic conflict of interest here with Musk owning most of both companies. And they're in entirely separate fields, there is no plausible synergy here to be gained.
Yeah but who can be hurt by this, these are both private companies? So whose interest is his "conflicting" with? I'm sure the shareholders will raise it with him and/or bring a lawsuit if they aren't happy (they probably are happy).
How can you have a conflict of interest if they're entirely separate fields? They have different interests, so where's the conflict?
You don't need synergies to justify a merger. They're often used as justification as in paying well above market price. But it has nothing to do with actual justification. You can just have a holding company of businesses
The conflict of interests here is the conflict between musks interests and the other shareholders interests
he is literally going to launch datacenters into space to train ai so they are a little related
edit: these replies aren't going to age well
Yeah, I'm not buying that. I don't see how that could be any cheaper than regular datacenters. It might just be technically feasible, but launching stuff into space will always be more expensive than not launching stuff into space. And all those pesky technical issues like cooling might be solvable, but I doubt they're that cheap to solve.
You're right, but in this sense:
literally (adverb)
informal : in effect : virtually
Used in an exaggerated way to emphasize a statement or description that is not literally true or possible.
Ex: I literally died of embarrassment.
He says he is going to launch data centers in space. We should all know better than to take him at his word on that by now.
No he is not. It makes no sense from a physics standpoint or an economic standpoint. And even if they were, it wouldn’t require whatever this acquisition is.
No, he's not. And if he does, he's as big of an idiot as his detractors say that he is.
Oh man, I sure hope he disclosed that
Musk has a history of having one of his more successful companies buying one of his less successful companies. xAI bought X, and Tesla bought SolarCity
Musk is notorious for shuffling assets across his companies to make some financials look better. For example, shuffling Twitter servers (and then all of Twitter) under xAI.
Apparently SpaceX is the ultimate vehicle for buying Elon's shit that no one else wants. They're also buying thousands of Cybertrucks.
https://futurism.com/advanced-transport/spacex-buying-unfath...
Need any solar shingles?
my partner got shingles a couple years ago, it was a very painful experience!
(to be crystal clear, I am making a joke equating the failed SolarCity/Tesla solar shingles to the (generally considered very painful) Herpes Zoster manifestation also called "shingles")
lol
Musk's involvement for one.
It's just "Elon zomg lulz" trolling for updoots. This place pretends to know better.
Well, at least this time both companies are fully private.
For the next four months, at least.
What's funny? Do you think the investors are against this? The investor's aren't idiots. I imagine the typical investor in Elon Musk's companies would approve of this sort of thing. So what's the problem? Besides, its a private company with Musk as majority shareholder in both. That's the beauty of private companies, you can just do things.
I wish more companies were private and ambitious. I'm tired of companies like Apple making marginal spec bumps to their phones and milking the same products for decades
> What's funny? Do you think the investors are against this? The investor's aren't idiots.
Any proof of that?
More than 70% of voting shares supported the package, very close to the level of support in the original 2018 vote. This excludes Musks share.
And consider that this is retroactive, meaning it's backpay. They're literally voting to give the guy $50b for work performed. He has a lot of confidence from his investors. And if there were issues, there would be lawsuits. Ironically the only lawsuits that get brought up, like the one about the pay package, are basically trolls, from a guy that had 9 shares.
Besides the parent is the one making a claim that something not above board is going on so burden of proof is on him.
Finally, it's a private company where Musk is the majority shareholder. He's moving money from one pocket into another, and any moves will be reflected in his attempt to raise money with the IPO coming this year.
Why do people online pretend not to understand?
Nothing in your argument is proof that the investors aren't idiots.
Apple just launched their own silicon chips just a few years ago. They're very ambitious but still calculated.
> > I'm tired of companies like Apple making marginal spec bumps to their phones and milking the same products for decades
At least what Apple does is real not make believe like everything Musk claims , disappear boring Apple or even boring Microsoft, Oracle, IBM etc.
And the world would come to a screeching halt, disappear all of Musk companies and people would barely notice.
You seem to be eager to be sold dreams , that's exactly what vaporware salesmen like Musk hope to find on their path
The investors want to cash out, Musk needs lots of money to plow into his latest toy that so far only excels in ridiculing him and sexual harassment/CSAM, so they make a deal to take in xAI and go public. Win win.
> The investor's aren't idiots
citation needed.
It's widely reported that Musk is a majority shareholder of xAI and the controlling shareholder of SpaceX (close to 80% of voting shares). Not surprising that he would be looking to consolidate ownership under one entity especially if he perceives significant synergies (i.e., data centres in space).
Data centers in space are a hilariously bad idea. Where would the heat go? This idea is like the opposite of liquid cooling.
Shocked to see SpaceX buy the datacenter in space meme. Where does the power come from? Where does the heat go? Why add (high) launch costs to your buildout capex? Why add radiation as another risk factor to your already-unreliable GPUs? Am I missing something fundamental here...?
Not to mention the huge issues of cosmic rays. Sure, if the lifespan of the satellite is expected to be low, then maybe tolerable. But even then, how would this be financially viable?
I didn't say it was a good idea, just that if Musk perceives it's a good idea then it makes sense why he would want to combine the two businesses.
I think it's far more likely that he wants to combine his businesses to roll his really expensive, debt-ridden companies into one entity with the company that actually reliably makes money.
Only a person who is high as a kite can think thats a good idea.
Aside from Elon Musk, there are a few other people with a lot of capital aiming to do the same thing. That means, either they are all wrong (possible) or this problem has been solved somehow and the solution itself is not public.
Google and Amazon are doing the same thing. Maybe it is a moonshot (pun intended), but Musk is hardly alone in the push.
https://www.wsj.com/tech/bezos-and-musk-race-to-bring-data-c...
https://www.nytimes.com/2026/01/01/technology/space-data-cen...
Some guy on hacker news argued they could just use radiators.
Radiators might be a reasonably effective way to reject heat if you can run your AI machines at 1000 K or thereabouts.
Indeed. But it's also a hilariously Musky idea! Some moderate technical competence paired with sociopathy and an ego orders of magnitude too big, and voila, you get Cybertrucks, Hyperloops, Neuralinks, Teslabots, datacenters in space, and all the other garbage the man spews.
I cannot wait for him to one day be hit in the face by reality.
I have never understood how Data centers in space ever make economic sense, the payload, latency and many other issues make it difficult at least for the immediate needs
Latency isn't an issue with Starlink - the data centers are in low earth orbit, not in GEO
You mean unlike Hyperloops, Cybertrucks, Teslabots, Neuralinks, and all the other insane stuff that moron cooks up?
This means that Grok, Elon's politically labotomized involuntary pornography generator, and X (formerly Twitter), Elon's Nazi-adjacent propaganda machine, are now completely intertwined with SpaceX, a too-big-to-fail government contractor that currently serves as America's only reliable option for manned orbital spaceflight.
Anyone who doesn't see how broken this situation is isn't paying attention. This is how people like Elon, who want to seize as much power from the government as they can, ensure that the means for seizing that power are untouchable.
Anyone who has ever used Grok or X lately knows that both of these products are heavily manipulated to align with the political, social, and economic views of Elon Musk, who is increasingly boosting "white power" language and full-throatedly backing America's most nationalistic and authoritarian president to date.
This is just another consolidation of power, and it's deeply worrisome. Any integrity one may have hoped remained at SpaceX just vanished when they aligned their mission with that of these deeply problematic digital services.
And this is not even scratching the surface of what looks like a deliberate attempt to create Kessler syndrome by launching millions of cheap short-term satellites into orbit, or the rationality of putting datacenters into orbit in the first place...
Seems more risky to me; it takes only one Mercurial temper swing for SpaceX to be nationalized, and now Twitter and xAI are bundled along with it.
I hate to be that guy that posts an AI-generated response, but I thought Grok had an interesting take on this problem:
Energy Generation Practicality
Building satellite datacenters powered by solar energy in space is conceptually feasible, as solar insolation in orbit is about 1,366 W/m²—roughly 35% higher than on Earth's surface due to no atmospheric filtering or day-night cycles in geostationary or sun-synchronous orbits. Advanced solar panels in space can achieve efficiencies of 25-35%, yielding 340-480 W/m² of electrical power. For context, a typical terrestrial datacenter rack consumes 8-15 kW, but AI/high-performance computing (HPC) racks can reach 30-80 kW. Scaling to a hypothetical 1 MW datacenter (enough for ~12-30 AI racks) would require approximately 2,000-3,000 m² of solar panels, assuming 25% efficiency and some losses.
However, practicality hinges on mass and deployment. Thin-film solar arrays (e.g., from projects like Caltech's Space Solar Power Project) can weigh as little as 0.5-1 kg/m², making the array for 1 MW ~1-3 tons. This is manageable for a constellation, but integrating with compute hardware adds complexity. Concepts like Google's Project Suncatcher and Thales Alenia Space's ASCEND study propose modular satellites with integrated solar power, potentially achieving 10x lower carbon emissions than Earth-based datacenters. Challenges include orbital shading in low Earth orbit (LEO) and panel degradation from radiation/micro-meteoroids, but overall, generating sufficient energy is practical with current tech if scaled modestly (e.g., 100-500 kW per satellite).
Heat Radiation Practicality
Heat dissipation in space relies solely on thermal radiation, as there's no atmosphere for convection or conduction to a medium. This is highly practical because space acts as an infinite cold sink (~3 K background temperature). Waste heat from datacenter compute (typically 80-90% of input power) can be radiated via large, lightweight panels with high emissivity coatings (ε ≈ 0.9).
Using the Stefan-Boltzmann law for blackbody radiation (P = εσA(T⁴ - T_sink⁴), where σ = 5.67 × 10⁻⁸ W/m²K⁴), a radiator at 50°C (323 K) emits ~550-620 W/m² (one-sided). For a 1 MW heat load, you'd need ~1,600-1,800 m² of radiator area—comparable to solar panel sizing and feasible with deployable structures like those on the ISS (which radiate up to 350 W/m²). Designs from Laird Technologies and NEC emphasize distributed radiation at the PCB level, using materials like aluminum or composites for efficient emission while minimizing solar absorption.
The main limitation is surface area: satellites must balance radiator size with stability and drag in LEO. Overheating risks exist during eclipses or high-load bursts, but passive systems (e.g., heat pipes) and active louvers make this solvable. Studies like ASCEND confirm space datacenters could reject heat more efficiently than Earth-based ones, avoiding water-intensive cooling.
Shielding from Cosmic Rays in LEO
In LEO (400-2,000 km), Earth's magnetic field deflects much of the high-energy radiation, resulting in dose rates of 100-10,000 rad(Si)/year—far lower than in higher orbits. This makes shielding practical and often minimal for short missions (1-5 years). Primary threats are total ionizing dose (TID) causing degradation and single-event effects (SEE) like bit flips or latch-ups.
Standard aluminum shielding (3-5 mm thick, ~1-2 g/cm² areal density) reduces TID to <10 krad(Si) over 3 years, sufficient for 65-130 nm chips. Polymers like polyethylene or borated polyethylene outperform aluminum by 20-70% in proton shielding, reducing displacement damage. Many LEO cubesats use commercial off-the-shelf (COTS) electronics with redundancy (e.g., multiple computers) and error-correcting code, avoiding heavy rad-hard parts that add mass/cost.
For datacenters, radiation-tolerant designs (e.g., from Melagen Labs' low-density composites) can extend COTS lifespan 3-5x. No exotic shielding is required—Earth's magnetosphere provides natural protection, and simulations (e.g., MCNP/OMERE) confirm viability. Drawbacks include occasional upsets during solar flares, but overall, LEO shielding is straightforward and cost-effective.
Launch Costs for Practical Mass and Competitive Compute
Launch costs are the biggest barrier. Current prices to LEO are ~$2,700/kg (Falcon 9) to ~$1,500/kg (Falcon Heavy). For a "useful" compute level—say, 1 petaFLOP/s (comparable to a small AI cluster)—hardware might mass 1-5 tons (based on dense racks like NVIDIA DGX at ~100-200 kg/rack, plus power/thermal systems). Including solar panels and radiators, a single satellite could total 5-10 tons.
At $1,500/kg, launching one unit costs $7.5-15 million—potentially competitive with terrestrial datacenter energy costs ($0.05-0.15/kWh over lifetime), but scaling to a constellation (10-100 satellites for global coverage) pushes totals to $75-1,500 million. Google's analysis requires costs <$200/kg by the mid-2030s for parity; Starship aims for $10-100/kg, dropping per-satellite costs to <$1-2 million.
Competitiveness: Terrestrial datacenters cost ~$10-20 million/MW to build/operate annually, but space avoids land/water constraints. At current costs, space is impractical for mass deployment; with Starship, it could undercut Earth prices by 10-50x in energy efficiency. Thales estimates needing 10x less emissive launchers for net-zero viability.
Overall, the concept is technically practical today for prototypes (e.g., via Starcloud or SpaceX proposals), but economically viable only with launch costs dropping 10-20x. Energy and heat are strengths; shielding is low-risk in LEO. Full constellations could emerge by 2030 if reusability scales.
People pooh-poohing space datacenters will obviously think this is a bad move. But Elon clearly believes space datacenters will work. Given that, and the fact that SpaceX will IPO this year, this acquisition was inevitable.
SpaceX and xAI would not be able to freely collaborate on space datacenters after the IPO because it would be self-dealing. SpaceX likes to be vertically integrated, so they wouldn't want to just be a contractor for OpenAI's or Anthropic's infrastructure. Merging before the IPO is the only way that SpaceX could remain vertically integrated as they build space datacenters.
How can you clearly believe anything Musk says at this point. It's not 'clear' at all what he actually believes and what he just makes up.
Space datacenters just so grok can undress women? this is the dumbest company on the planet