Nobody Needs Starship*

Who needs super-heavy reusable lift, not as a vision, but as a line item on a program of record? Following the money of the largest IPO on record and analyzing demand for the world's largest launch.

Men lie, women lie, numbers don’t.

—Jay-Z

The following is neither financial nor legal advice, but solely the opinion of the author. It’s still highly entertaining and informative, according to proof-readers. The author holds no position, long or short, in SPCX or related instruments.

Yes, the AI article is coming. But with the SpaceX IPO looming over the horizon like a swinging sword of Damocles (or like data centers in low-Earth orbit (LEO)), Germany allocating more and more of its €35 billion space defense budget, and Russia-v-Ukraine experiencing a tilt since Elon turned the taps off for Vladimir (by which I mean that SpaceX stopped Russia’s “unauthorized use” of Starlink), this needs to be addressed NOW, not later.

But first, a clarification. This article is not about the SpaceX IPO. Well, in the end it kind of is, because SpaceX’s recent numerology exercises (financial engineering?) have provided me with the final puzzle pieces for writing this, and because their commercial activities are key to their defense offerings (I’ll explain below). This article is about a super-heavy reusable launcher system, and whether humanity has any need for it today, tomorrow, or in the future.

The business case for Starship is deeply intertwined with the business case for SpaceX, and that matters for the question of whether a Starship-like super-heavy launch vehicle is viable at all. The reason is two-fold: 1) commercial and civilian demand acts as a kind of subsidy (other than actual subsidies), and 2) without sufficient commercial demand, there might not be a reason for building and maintaining the launch and turnaround/maintenance infrastructure such a system needs. A launcher in this class only makes economic sense at high cadence, and high cadence only makes sense if somebody is buying the flights.

On 22 May 2026, Starship flew its twelfth integrated test flight, the first of the new V3 generation. During boostback, engine E6 on Booster 19 exploded and took neighboring engines with it, and the booster hit the water hard. The ship continued, reached its targets, and released 20 Starlink mass simulators plus two modified V2 satellites. Musk called it “an epic first Starship V3 launch”. Both halves of that flight are the program in miniature: the most capable launch vehicle ever built (worth saying clearly before taking it apart), and a program that now stands at seven successes in twelve attempts; counting Flight 12 as a success because the ship completed its objectives. Count every lost vehicle as a failure, however ‘epic’, and the score drops to six, or five.

So the question this article asks is not whether Starship is impressive. It is. The question is who needs super-heavy reusable lift as a line item on a program of record, with a budget line, a payload manifest, and a recurring requirement behind it. The method: follow the money through SpaceX’s S-1 filing, through China’s megaconstellation cadence, through Europe’s launch demand, through the physics of the orbital data center, and through the one customer profile that actually fits.

What You Will Learn In This Deep Dive

1. Commercial vs. Defense?

2. Europe’s Slump

3. Comms

4. Data Centers

5. Mars

6. Moon

7. Who Pushes the Space Data Center “Case” and Why

8. Why Starship Exists

9. The Future of Starlink

10. The Future of Launch

11. A Potential Wildcard

12. A Potential, Scary, Out

13. David Blaine Economics

14. Recommendations for MoDs, Investors and Industry

15. Wrap-up: Nobody Needs Starship, *Except…

The TECH WARS Deep Dive you’re about to read: 37 pages of condensed insights, many of which you will find nowhere else.

Commercial vs. Defense?

This is the first big misunderstanding. Outside of Europe, there hardly is such a thing as commercial vs. defense. It’s commercial and defense. I’m lucky to have worked on both sides of the Atlantic, which is the best way to learn first-hand that the EU’s timid separation of defense and commercial makes little sense. Both go together, hence why it’s called dual use. A great example is the United States’ Hybrid Space Architecture (HSA), but I talk a lot about US space defense plans on this publication (I even compared Germany’s plans to the ones of the US), so let’s talk about China for once, because we still don’t look at China enough in the West, despite its massive progress. Three major projects are running in the Middle Kingdom:

1. Guowang

2. Qianfan

3. Honghu-3

Guowang (national network) is the state-run answer to Starlink, operated by China SatNet under CASC. Its ITU filings cover 12.992 satellites across two sub-constellations, one shell at 500 to 600 kilometers and one at approximately 1.145 kilometers. As of early June 2026, roughly 168 satellites are operational after about twenty launch missions. The deployment plan, according to Chinese industry reporting, calls for 310 satellites in 2026, 900 in 2027, and 3.600 per year from 2028. Researchers at the PLA’s Space Engineering University have published openly on Guowang’s role in countering Starlink and securing orbital resources, and there have been recurring reports, unconfirmed by independent sources, of the constellation being tested for missile-targeting applications near Taiwan. Whatever broadband Guowang ends up selling, the program’s framing is military-first.

Qianfan (Thousand Sails, also marketed as Spacesail, formerly G60), backed by the Shanghai municipal government and operated by Shanghai Spacecom Satellite Technology, reached 200 satellites in orbit on 5 June 2026 after a Long March 8 launch from Wenchang. Phase 1 targets 1.296 satellites; the full filing covers up to 15.000. Each satellite is a flat-panel design with a mass of approximately 300 kilograms and a krypton Hall-effect thruster. Qianfan is the commercial face of the pair: trial-service agreements have been reported in Brazil, Malaysia, Kazakhstan and Turkey, and it’s the instrument with which China is contesting the Global South broadband market, plus the spectrum and orbital slots that come with it. Worth noting: deployment was paused for months in 2025 after reported thruster and gyroscope failures on orbit, resuming in April 2026. China’s flagship commercial constellation is not flying flawlessly, either.

Honghu-3 is the third filing, a planned constellation of approximately 10.000 satellites by Hongqing Technology, in which LandSpace holds a stake of approximately 48 percent. That is the same LandSpace that builds the Zhuque-3 reusable rocket, which tells you how Chinese launch and constellation interests interlock. Yes, China has reusable rocket companies, mostly building, not yet operating reusable rockets. But it has more than a dozen of them. Dear Europe… we’ll need to talk about you (two paragraphs below this one).

Now, let’s look at the launch side of this, because it’s the part that matters for our quest(ion). China launched 92 rockets in 2025, a national record, and 37 launches by June 5th, 2026, and is targeting more than 100 for the year. Every single one of those constellation satellites rides on expendable Long March 6A, 8, 8A, 12 and 5B vehicles, nine to eighteen satellites at a time. The reusable fleet is still in testing: LandSpace’s Zhuque-3 reached orbit on its maiden flight on December 3rd, 2025, and lost the booster during the landing attempt (per LandSpace, an anomaly after first-stage reignition in the landing phase), and the state’s Long March 10A flew a suborbital booster-recovery test on February 11th, 2026, with a controlled splashdown next to a sea platform. In short: the country with the most aggressive satellite deployment plans on Earth is executing them right now with no super-heavy and no reusability, at a cadence of two launches a week. Keep that in mind for later.

Europe’s Slump

Europe has huge mental, moral and arithmetic issues with this idea, still, no matter what you read about the European Defence Agency and the European Space Agency (EDA and ESA, respectively) working on projects together. Talk to the most senior Trekkie-, Apollo-Generation space execs in the US, and they’ll happily admit that they’ve got projects with the Defense Innovation Unit or Department of War (DIU, DoW), even if their main designated market is civilian-commercial. Try to have the same conversation with any European equivalent, and the air pressure drops rapidly: startups can’t because the limited partners (LPs) of their VCs exclude defense, because they’re still banking with a bank which has their ESG-tail between their legs, or their main customers would shut the faucet the moment they hear that the company is serving defense customers as well. Airbus and Airbus Defence and Space are not only two separate entities but two separate brands (vs., e.g., Boeing), and I’ll stop there.

What does this produce on the launch side? Numbers, so let’s look at them. Ariane 6, Europe’s flagship launcher, first flew in July 2024 after years of delay and had logged four successful flights in its first sixteen months. Four. The target was six flights in 2025 and eight in 2026; ArianeGroup leadership has said publicly that a cadence of ten will only arrive with IRIS² deployment toward 2029, which is its own can of worms/false assumption that a low number gets you a resilient constellation. The first four-booster Ariane 64, the variant that Amazon’s constellation and IRIS² actually require, flew in February 2026. The backlog is real: more than 30 missions, 18 of them for Amazon. Pierre Lionnet of Eurospace has argued that Ariane 6 reflects the market projections of ten or twenty years ago, and that demand alone will not carry the launcher to the cadence its cost model assumes. That is a polite way of describing a rocket sized for a market that moved on.

Isar Aerospace

The startup layer is more hopeful. Under the European Launcher Challenge, member states pledged 902,2 million Euros across five companies: Isar Aerospace and Rocket Factory Augsburg from Germany, MaiaSpace from France, PLD Space from Spain, and Orbex from the UK (didn’t make it). Isar’s Spectrum failed on its first flight in 2025, and a second attempt was in preparation at Andøya as of January 2026, then moved, and finally scrapped in April 2026. No matter the setbacks (it’s space and space is hard), one development is crucial: These are small and medium launchers, and that’s the appropriate size, because here’s what European institutional demand actually looks like:

IRIS², the EU’s sovereign secure-connectivity constellation, consists of approximately 290 satellites: 264 in LEO at approximately 1.200 kilometers and 18 in MEO. It costs 10,6 billion Euros under a twelve-year concession to the SpaceRISE consortium (SES, Eutelsat and Hispasat, with Airbus Defence and Space, Thales Alenia Space, OHB and others as suppliers), with a first launch planned for 2029 and service from 2030, on a total of 13 Ariane 64 flights. Thirteen launches over roughly two years. That’s the launch demand created by the sovereign connectivity program of a continent of 450 million people. Let that sink in.

And Germany? On 25 September 2025, Defense Minister Boris Pistorius announced at the BDI Space Congress in Berlin that Germany would invest 35 billion Euros in space-related defense projects by 2030, a sum in the region of ESA’s entire budget over the same period. Five priorities: hardening against jamming and cyberattack, space situational awareness, redundancy through multiple networked constellations (which, I have argued, are not proliferated enough and missing a sustainable replenishment plan), secure and on-demand launch, and a military satellite operations center under a new Bundeswehr Space Command. Pistorius named the Russian Luch-Olymp inspector satellites that have been shadowing Bundeswehr-leased communications satellites as strategic triggers, and, a genuine doctrinal first for Germany, put offensive options on the table. I have analyzed the architecture behind this in Germany’s Rise as a Space Power. For our question, the core point is this: Per the Bundeswehr’s own communication, heavy-lift requirements will continue to be met by Ariane 6, and the domestic providers named are RFA, Isar and HyImpulse, all in the small-launch class. 35 billion Euros of new European space defense money, and not one Euro of it requires a 100-tonne fairing.

Back to Starship and space that actually moves forward. Because it actually does.

Hi! Glad You’re Here

If you’re subscribed to TECH WARS, you are likely one of the few defense professionals who demand all of the dots to be connected, from technologies to training, to tactics, operations, strategies, logistics, defense economics, force structure and posture. I can imagine that you’re the kind of person who often asks themselves questions like “Could this tactical-technological iteration accelerate into a strategic effect?” If this is you, then thank you, I’m glad that you’re here. You are part of a very small, open-minded group who aren’t scared to question defense paradigms and prevailing currents. Thank you for subscribing to my research, analyses, and ramblings. If you have a colleague who thinks like you, or who would benefit from some perspective, would you mind letting them know about TECH WARS? I know you don’t need it, but I’ll give you free months for it. Thanks for reading TECH WARS and thanks for spreading the word!

Thank you for supporting my work!