Discussion

Why the world's militaries are scrambling to create their own Starlink

kolinko: The thing is - without Falcon9 / Starship they really cannot - both China and EU are ~10-20 years (sic) behind SpaceX, and without thousands of satellites on LEO you just cannot have terminal similar to SpaceX's.(And don't get me started on how bad Iris2 is/will be. It's a program that EU has to shut down discussions on how terribly behind we are.The last time I checked, a year ago, EU's plans were to have first Falcon9-level flights around 2035 (!!!), and that was assuming no delays, so absurdly optimistic. Adding a few years for ramping up the production, 2040 is the earliest we can have optimistically something like Starlink from 2020.

db48x: SpaceX will happily launch satellites for competitors. OneWeb has bought launches from them, for example.

bryanlarsen: Or at least they were while anti-trust still had some teeth. Trump's DOJ is highly unlikely to go after Starlink for refusing to launch for a competitor, let alone another nation's military.

thisislife2: Can you explain what makes Falcon9 / Starship special (or needed) to launch these satellites? China, India, EU, Japan etc. all have the capability to launch satellites. So why is a Falcon9 / Starship a particular requirement?

mooreds: Cost, maybe? It is one thing to ship up a valuable satellite (which they all can do). But to ship up 1000s of satellites (and keep doing it in perpetuity, because IIRC they don't have a long lifetime[0]) gets expensive.0: Looks like 5 years. https://www.space.com/spacex-starlink-satellites.html

tartuffe78: Starlink is apparently 65% of all active satellites, it would be very expensive to emulate that without super efficient launching capabilities.

samrus: Has to be the cost. A reusable launch vehicle is such a ridiculously better value proposition that it creates a discrete evolution. Some things just arent feasible to do without them

_whiteCaps_: In Canada, the CF is working on rebuilding their expertise in HF radio, as they realized that in case of large scale conflict, satellite systems aren't going to be dependable.https://en.wikipedia.org/wiki/Canadian_Forces_Affiliate_Radi...

Bender: Starlink's first customer was supposed to be the US Army. I am curious what requirements they did not meet.

bryanlarsen: Falcon-9 first landed in 2015 and was regularly landing within a couple of years. So being 10 years behind means "almost ready to go".suborbital Yuanxingzhe-1 landed may 2025, and orbital Zhuque-3 was really close to landing in December. Long March 12A also tried in December although it wasn't as close to success.So if China is 10 years behind, they've caught up. We won't know if they're 10 years or further behind for a couple years more, though.And while China may be 10-15 years behind on their Falcon-9 equivalents, they're likely less than 10 years behind on their Starship equivalents.

jmyeet: The story I like to tell is about the Manhattan Project. This caused a debate in US strategic circles that set policy for the entire post-1945 world. Debate included whether a preemptive nuclear strike on the USSR was necessary or even just a good idea.Anyway, many in these circles thought the USSR would take 20 years to develop the bomb if they ever did. It took 4 years. The hydrogen bomb? The USA tested theirs in 1952. The USSR? 1953.China now has decades of commitment to long-term projects, an interest in national security and creating an virtuous circle for various industries.The US banned the export of EUV lithography machiens to China but (IMHO) they made a huge mistake by also banning the best chips. Why was this a mistake? Because it created a captive market for Chinese-made chips.The Soviet atomic project was helped by espionage and ideology (ie some people believed in the communist project or simply thought it a bad idea that only the US had nuclear weapons). That's just not necessary today. You simply throw some money at a few key researchers and engineers who worked at ASML and you catch up to EUV real fast. I said a couple of years ago China would develop their own EUV processes because they don't want the US to have that control over them. It's a matter of national security. China seems to be 3-5 years away on conservative estimates.More evidence of this is China not wanting to import NVidia chips despite the ban being lifted [1].China has the same attitude to having its own launch capability. They've already started testing their own reusable rockets [2]. China has the industrial ecosystem to make everything that goes into a rocket, a captive market for Chinese launches (particularly the Chinese government and military) and the track record to pull this off.And guess what? China can hire former SpaceX engineers too.I predict in 5 years these comments doubting China's space ambitions will be instead "well of course that was going to happen".[1]: https://www.theinformation.com/articles/china-want-buy-nvidi...[2]: https://www.scientificamerican.com/article/chinas-explosive-...

zitterbewegung: To be future proof for more administrations you don't want a monopoly at any step. you really want at least three competitors at minimum. Large companies in tech have realized this by now since the 90s. Recently TeraWave was launched by SpaceX due to the inherent risk (and this is a direct competitor to SpaceX. See https://www.cnbc.com/2026/01/21/bezos-blue-origin-satellite-...

spwa4: Ever notice just how many countries seem to be pretty convinced war is coming? And don't tell me it's all Trump, at the very least they believe that whoever follows Trump isn't going to be very different. Plus it's mostly EU that's rearming, and surely they aren't afraid they'll be attacked ...

Joel_Mckay: Canada has a lot of obscure technology that would normally fall under export restriction in the US.The problem I have with the Canadian business culture was there is zero protection on a global scale for your company, privacy, and or personal safety. =3

roughly: EU had a reliable military and technological partner in the US until circa 2016, and maintaining that belief became untenable in 2024. The reason EU countries are all of the sudden investing in onshoring critical military capabilities is that until Trump it’s been the policy position of the US to prevent them from doing so by doing it for them, a policy we inaugurated after WW2 and expanded during the Cold War for various reasons that we seem very sure don’t apply anymore.

spwa4: I've worked in defense tech. This is true, but it should be described much more as "Europe believed US would save their ass - for free, and did nothing" (with exceptions, like France, and some token efforts within NATO) The US was not holding back much within NATO.

ciupicri: > many in these circles thought the USSR would take 20 years to develop the bomb if they ever did. It took 4 years.Because some people committed treason and gave the technology to the Soviets.

jordanb: I think the next big war will involve a kessler syndrome, not because people start firing off anti-satellite weapons (since there's a strong component of MAD in doing that) but because the belligerents will have their own multi-thousand satellite constellations in orbit and they will quit coordinating with one another on collision avoidance.

childintime: A smaller player like North Korea and Iran would not have as much to lose. Iran is doing something similar today, suicide bombing everything it can.

josefritzishere: It's worth pointing out that aside from Elons behavior the real issue with Starlink is that it's insolvent. Starlink does not make money. (The solvency gap is hotly debated) But that fact means it's long-term reliability is in question. No military wants to risk that kind of system dependency.

tekla: None of those countries (well probably except China) have any significant launch capacity to deploy constellations

bluGill: They can build it in a few years though. It takes money and can be done overnight but there is nothing about that that costs 10 years. 10 years got to the moon - from a much lower base. 10 years means you are starting with college graduates and building it from no previous experience - or you already have a lot but only are putting minimal budget into improving.

bryanlarsen: Starlink is redeploying to 300 miles. Many consider Kessler to be impossible at 300 miles. Any unpowered satellite at a 300 mile orbit will deorbit within a couple of months. But a collision means fragments which deorbit faster because they have a higher surface/weight ratio, and because orbit disturbances lower that time considerably. Any single disturbance that raises aphelion lowers perihelion.

Razengan: God can we have an alien invasion already PLEASE12 000 years of this shit

sigmoid10: China also had made industry espionage their way to go in these things. They are not even hiding it anymore. It's almost comical how much they copied SpaceX. And I'd be surprised if they hadn't supply-chained themselves into some level of access in all the big aerospace corpos by now. But Europe? Developing this kind of stuff from scratch in a few years without an unregulated messy startup ecosystem and no army of state sponsored hackers? No chance.

palmotea: > Because some people committed treason and gave the technology to the Soviets.American big business is pretty much doing that every day, handing over technology to increase China's manufacturing tech level.

palmotea: > Starlink is apparently 65% of all active satellites, it would be very expensive to emulate that without super efficient launching capabilities.But does a military really need that many to get the necessary capability? Would a smaller constellation be sufficient, especially without competing civilian users?

icegreentea2: I'd broadly agree that EU is pretty behind the curve. But I think China is probably only ~5 years max behind the curve in terms of Starlink.But in terms of defense needs, I don't think you actually need the thousands and thousands for reasonable returns. DoD/NRO has bought maybe ~500 Starshields (https://www.fool.com/investing/2024/03/26/spacex-starshield-...) from SpaceX.I think China is well within reach of being able to put up those numbers within a few years, even if they don't get re-use figured out (which I think they will within a 2-3 years - basically what SpaceX did from the first landing attempts to success).

bluGill: Sorry, relativity is against it. They - if they exist (a debate I'm not touching) - don't even know we are here. Even if they knew we are here they can't get here.

Razengan: Or have a hands-off policy like we do with uncontacted tribes and some protected animal populations etc.I'll settle for anything to be honest. A sign, a derelict, an artifact, a fossil, an echo.. anything to distract humans from shitting on each other for a little while at least.

bluGill: Again, physics says they can't. relativite and signal degrigation is hard. the energy of a star outside our arm of the galaxy isn't easy to detect, much less any signal of lower power.

vardump: Anything to back that up? Starlink is widely considered profitable.

jmyeet: There's a deeper message here. I believe that countries around the world are moving towards a stance that the US is an unreliable partner and that their national security depends on not being reliant upon the US.An obvious place for this is that I think the EU will follow China's stance on not wanting to be beholden to US tech companies. The EU will bootstrap this by requiring EU government services to be hosted on platforms run by EU companies subject to EU jurisdiction. Think EU AWS. This is easier said than done.But this is really a consequence of the current administration having absolutely no idea what they're doing and they're intentionally and unintentionally destroying American soft power.Another way this can come to pass is that the EU decides that the US is an unreliable partner for their security needs so you will find that various weapons, vehicles, platforms, etc for EU militaries will be supplied by local companies, particularly if the US effectively abandons Ukraine.Starlink is just another piece of that.The current administration paints NATO as Europe taking advantage of the US. It could not be more wrong. NATO is a protection racket for the US to sell weapons and control European foreign policy.We kind of saw a precursor to all this with GPS. For anyone who has been around long enough, GPS used to be less accurate, deliberately. Why? Because defence (apparently). There was a special signal, Selective Ability ("SA") [1], that military gear could decode to be more accurate.Fun fact: one of the clues to the first Gulf War was that the military turned off SA on the commercial GPS system because they couldn't procure enough military equipment so had to use civilian gear [2].I think Europe was slow to learn the lesson of being completely reliant on the US but we did end up with Glonass and Galileo as a result.To exert the kind of control the US does through tech platfoorms, the US needs to be predictable and reliable can't be too overt with exerting political influence such that American imperial subjects can pretend they're still independent. This administration has shattered that illusion.[1]: https://www.gps.gov/selective-availability[2]: https://www.spirent.com/blogs/selective-availability-a-bad-m...

palmotea: > There's a deeper message here. I believe that countries around the world are moving towards a stance that the US is an unreliable partner and that their national security depends on not being reliant upon the US.That's not a bad thing, because the EU has been a mooch since the end of the Cold War, at least. It's unfortunate it took two terms of Trump for them to finally chance their attitude.

SlinkyOnStairs: Another major detail is that SpaceX is simply burning enormous amounts of money on this.Starlink's revenue is comparable to the ESA's entire 5 billion euro budget, and it still looks like starlink is not net-profitable as a service. (And kessler syndrome avoidance is already pushing up costs with the lower orbits)The chief problem "stopping" other countries from developing a starlink competitor is that starlink simply doesn't make all that much sense if your country is capable of basic infrastructure construction. Fiber runs are expensive but not that expensive.

JumpCrisscross: > it still looks like starlink is not net-profitable as a serviceStarlink was profitable in 2024 [1] and should be materially profitable once V3 goes up.> kessler syndrome avoidance is already pushing up costs with the lower orbitsThis hits everyone. And it’s not a serious cost issue. Starlinks are still being deorbited before they need to be due to obselescence. And the propellant depots SpaceX is building for NASA tie in neatly if the chips stablise enough to permit longer-lasting birds.> doesn't make all that much sense if your country is capable of basic infrastructure constructionInfrastructure gets blown up and shut off. Hence the military interest.[1] https://www.pcmag.com/news/how-much-does-starlink-make-this-...

SlinkyOnStairs: > Starlink was profitable in 2024Those are revenue figures.> This hits everyone. And it’s not a serious cost issue.That it affects everyone just makes the problem worse. If China or the EU does commit to a starlink competitor, there's even more crowding in orbit. Even more collision avoidance required.> Starlinks are still being deorbited before they need to be due to obselescenceThat's the point. These things are not staying up long, and they're staying up shorter and shorter.The constellation is both expensive to build and to maintain. That makes it a lot of trouble compared to running a bunch of fiber once and having only occasional maintenance trouble when some idiot drags a backhoe through it.> Infrastructure gets blown up and shut off. Hence the military interest.The military interest is real, but it remains to be seen how much money they're willing to put up for it. Higher latency more conventional satellite internet will have significant cost savings in comparison.

JumpCrisscross: > Those are revenue figuresAnd also net income.> just makes the problem worseDid you skip the part where it’s not a serious cost issue? None of these birds are even close to being propellant restricted.> These things are not staying up long, and they're staying up shorter and shorterBecause they’re being intentionally deorbited to make room for better birds. They don’t have to be deorbited as quickly as they are. But overwhelming demand makes it a profitable bet.

toomuchtodo: What would the cost be to deny these orbital altitudes?

SlinkyOnStairs: Incalculable.The cost isn't in paying someone to not use the orbit, it's that the busier a part of space gets, the more expensive it becomes to do collision avoidance and station keeping.What makes this impossible to calculate is that there's an unknown exponential involved. The more satellites, the more collisions that need avoiding. And the higher the chance that one avoidance will create new future collisions.At some point the space is simply so busy that collisions can no longer be avoided.

tristor: All of that, and the funny thing is /that is the easy part/. Moving payloads to space is just incredibly expensive, but not fundamentally hard in the same way that post-launch coordination of satellite constellations and RF tuning to support things like mobile connectivity are (I can connect to Starlink satellites from my iPhone through T-Mobile).