Launch itself isn't super profitable and doesn't have a very big TAM, especially if you carefully treat the TAM as a present value and discount future cash flows. Also, at least by my personal guesswork, Starlink alone shouldn't get close to a $1 trillion valuation. It seems like the current valuation is based on people's guesses about the likelihood that space data centers pan out in a big way.
The Starship program is one of - if not the - biggest risks here. Making starship work could bring the launch cost down far enough to derisk that whole program, and today just made it a lot less risky
You do understand that a company needs to have a business case equivilient to its valuation right?
Going to Moon only costs money, going to Mars only costs money.
Our whole planet sends 1000-2000t of payload into space, most of it is now starlink.
The market values SpaceX over multiple areas of the business and their future plans and it isn't just this event which they have proven to accomplish.
> You do understand that a company needs to have a business case equivilient to its valuation right?
Nope. That is not how this works nor how investors see it. It is about their total addressable market (TAM) which SpaceX can capture over everyone else. So here are some more further questions:
1) Who is paying SpaceX for the compute that they have?
2) How quick is SpaceX building their data centers to support their customers?
3) How many people use Starlink as their primary internet provider?
You understand that almost all of where SpaceX businesses can expand into those areas and grow into that valuation and higher?
> Our whole planet sends 1000-2000t of payload into space, most of it is now starlink.
Your point is?
So who else can catch and reuse a rocketship to frequently deploy satellites into space in a very short period of time? (Don't say Blue Origin.)
2) ? I don't understand what you mean. Are we now talking about Datacenters for others or the space thing?
3) probably below 8 million. 10 Million customer overall, plenty as an alternative or for camping etc.
Amazon Leo is an alternative. But for sure Space-X started a race to the bottom (china, blue origin and co) so they have to start moving and use their first mover advantage.
My point is, that Space-X is already responsible for the most amount of payload which gets send into space. We as a whole planet don't have that much payload and Space-X made it alot cheaper to send payload up.
So were is the 1 Trillion evaluation? Btw. its currently at 1.89T
The Grok models and products also use the compute in SpaceX's data centers. It only makes sense to provide that excess compute to others such as Anthropic that don't own any, and that is worth tens of billions a quarter.
> In parallel Neoclouds are building up the same compute and its more like a negative sign that they got lucky
It is not the same. You can ask any chatbot, how many GPUs do the neoclouds have over SpaceX in their data centers?
Then:
> 2) ? I don't understand what you mean. Are we now talking about Datacenters for others or the space thing?
Both. The point being that they (SpaceX) are already quickly building new ones on land codenamed "Gun Range" and "Spam" and are the only ones that can launch and connect all of them via Starlink (if they eventually launch new ones into space) and serve both themselves and new customers.
> Amazon Leo is an alternative. But for sure Space-X started a race to the bottom (china, blue origin and co) so they have to start moving and use their first mover advantage.
Amazon Leo is not even generally available. So that doesn't even count.
> My point is, that Space-X is already responsible for the most amount of payload which gets send into space. We as a whole planet don't have that much payload and Space-X made it alot cheaper to send payload up.
So what happens when the neoclouds are not able to build data centers on land anymore due to legislation that prohibits more of them getting built [0] and ultimately they get paused? Most of the neoclouds are already sold out and cannot build more of them and have deadlines on these loans with interest on them.
Maybe the compute has to be deployed somewhere else in the long term....
> So were is the 1 Trillion evaluation? Btw. its currently at 1.89T
Exactly. It if goes lower than 1T it is cheap. This is not even pricing in the building of their own semiconductor fabricator (Terafab).
[0] https://www.theguardian.com/us-news/2026/nov/04/datacenters-...
This DC can then poisen our atmopshwere btw. and burns up limited resources on our own planet.
They don't need the compute because their product Grok is not competitive. They don't use the compute to make Grok better because?!
Amazon Leo will be though. Space-X has a first mover advantage, what did Tesla do with it?
This isn't a massive problem for SpaceX as you want to believe it to be.
> This DC can then poisen our atmopshwere btw. and burns up limited resources on our own planet
So are you going to tell them to pause everything? (You won't) What solution do you have to them to solve this? (You don't have any solution.)
> They don't need the compute because their product Grok is not competitive.
That is false. Otherwise they would have given up on scaling Grok already.
> They don't use the compute to make Grok better because?!
So Cursor does not have any customers that use Grok on a regular basis beyond Twitter / X? Nor does Grok Bot need the compute either? (They most certainly do.)
Efficiency gains in their inference engines allow them to use less GPUs for themselves, and also sell that excess compute to others who are otherwise sold out. Both scenarios can be true and it doesn't mean they "don't need it" either.
> Amazon Leo will be though. Space-X has a first mover advantage
With what launch pad? SpaceX has plenty of their own and are building even more.
> What did Tesla do with it?
They built self-driving cars with "FSD" and autopilot with that advantage, and your point is?
It was meant to be a dual catch of the booster and the rocketship and only the rocketship itself was caught.
Catching both of them in another mission would top this one.