Welcome to The OuterSpark, here I make sense of the space industry one week at a time: what's being built, who's building it, and how you get in!
Quick housekeeping:
If this landed in your Promotions or Spam folder, drag it into your main inbox. Replying works too.
Make sure to subscribe so the next one lands directly in your inbox.
AD: I run The OuterSpark on Beehiiv, and setting it up has been the easiest thing. If you've been sitting on the idea of starting your own newsletter, you can set one up here.
I have been thinking about this for the past two weeks, and I still do not have a clean answer. So this week you are getting my working thought process rather than a neat conclusion.
Roughly $700 million went into space data centers this summer. On top of that, SpaceX committed $100 billion to ground infrastructure to launch them. I think putting computers in orbit is the right move. I also think it might quietly wreck the ozone layer, and almost nobody covering the funding rounds is talking about it.
Both things are true at once and that is the problem!
The $700 million orbital compute boom
Starcloud raised $250 million in August, doubling its valuation to $2.3 billion! They already have an Nvidia H100 GPU sitting in orbit right now, training AI models. Now they want permission for 88,000 spacecraft.
Fig. Cowboy Space logo. Source: Website.
Cowboy Space raised $275 million at a $2 billion valuation. Founded by Robinhood co-founder Baiju Bhatt, their approach is clever. Instead of launching a satellite that turns the rocket stage into space junk, they turn the upper stage itself into the data center. One vehicle, one megawatt of compute, launching in 2028.
Fig. Starbase Louisiana. Source: SpaceX.
Then SpaceX announced Starbase Louisiana. That is a $100 billion commitment for ten launch pads and up to thirty Starship flights a day, largely to lift their own orbital data center constellation.
Why the sudden rush? Because Earth's power grid has hit a wall. In major US markets, getting a new data center connected takes five to seven years. You cannot code your way around a seven-year wait.
Climbing the Kardashev energy ladder
To see why space makes sense, look at the Kardashev scale. In the 1960s, Soviet astronomer Nikolai Kardashev ranked civilisations by how much energy they can harness:
Type I: Uses all energy on its home planet.
Type II: Captures the full output of its star.
Type III: Harnesses energy across an entire galaxy.
Fig. Kardashev Scale visualisation.
Humanity sits around 0.7. We are basically stuck.
Earthbound data centers chew through massive amounts of power, water, and land. Every gigawatt given to an AI data center is a gigawatt taken from a city or hospital.
In orbit, those problems vanish. Solar energy is constant. Cooling happens naturally by radiating heat into the vacuum of space. No water, no land disputes, no grid wait times. It is our most realistic path to scale without consuming our planet.
The hidden ozone risk
Here is what keeps me awake.
Starcloud filed for 88,000 satellites. SpaceX wants up to a million. China planned over 51,000 across its Guowang and Qianfan networks. Over the next fifteen years, we plan to launch 100 times more satellites than humanity has launched in total history.
Everyone talks about space debris, but the bigger issue happens when these satellites come down.
Satellites are built mostly of aluminium. When they burn up during reentry, that aluminium turns into alumina particles. Alumina does not just sit there. It acts as a catalyst, speeding up the chemical reactions that destroy ozone molecules.
A 2024 study in Geophysical Research Letters showed that 2022 satellite reentries raised atmospheric aluminium levels by nearly 30%. If these mega-constellations launch, annual alumina pollution will jump by 650%, adding 360 tonnes every year.
Fig. Visualisation of space debris poluting the atmosphere and speeding up ozone layer breakdown.
In February, researchers in Germany physically measured this pollution plume in the sky. It takes decades for these particles to settle into the ozone layer. What we launch today presents a bill in the 2050s.
We spent forty years fixing the ozone layer. Accidental damage while trying to save Earth's power grid would be an absurd way to fail.
A simple rule change
The answer is not banning commercial launches. That leaves us stuck with slow state programs and an overloaded grid.
Instead, regulators should set deployment rate limits.
When you limit launch speed, engineers adapt. Right now, nobody gets paid to build satellites that last fifteen years, or vehicles that can be refueled and repaired in orbit, or hardware made from non-toxic materials. Cheap disposable hardware wins because launch costs are low.
Strict limits force better engineering. Emissions rules made cars better. Deployment limits will do the same for spacecraft.
We need to fix the regulatory rules before the rockets fly!
Upcoming Deadlines & Events
If you're working on getting into space, or building something in it, here are some upcoming deadlines that you should know of:
Deadlines
UK: UK Space Agency C-LEO Call 3, Stage One closes midday Monday 7 September. £42m for satellite comms tech.
Europe: ESA Future Navigation Demonstrators, open call. Ideas in before 30 September for the next selection round.
US: NSF America's Seed Fund, full proposals 4 November. You need an invited Project Pitch first, which takes one to two months, so pitch now.
US: NASA SBIR/STTR, rolling topic batches through to September 2027. Sign up for alerts.
India: IN-SPACe Seed Fund (up to ₹1 crore) and Technology Adoption Fund. Open in batches, so register on the portal and watch.
Events
UK Earth Observation Conference : 15 to 17 September, University of Warwick
UK Space Power Conference: 23 and 24 September, IET Savoy Place, London
If you think I called this wrong, hit reply and tell me why.
See you next week.
Ricky
