Your grandmother thinks data centers are just big warehouses with computers.
Your government thinks building more of them solves the AI infrastructure problem.
Both are about to be catastrophically wrong.
A week ago, Kevin sent me an article with six words that basically detonated my brain:
“Google. Space-based data centers. Read this.”
I opened it half-curious, half-exhausted, the way you doomscroll Instagram after a mentally draining day ,where your attention spans feels like a fragile, flickering candle. Five minutes later, my eyes widened in the same way they did when I first learned Elon Musk planned to land rockets vertically (and then did it).
I messaged Kevin:
“Wait… is this real or am I spacing out?” Ha, geddit?
Nope. Real. And if the math holds, this might be the most important infrastructure shift of the next decade , do you SEA it (another one! )
Why Google Wants to Move Compute to Space (and Why It’s Not Crazy)
Google’s Project Suncatcher is exactly what it sounds like:
AI data centers… in orbit.
Not in 2080. Not when we’re all old and retired.
But plausibly in the next 10–12 years. (Boom!)
Why space? Because space quietly solves the three things governments in Southeast Asia lose sleep over (EVERYBODY NEEDS SPACE,duh) :
1. Unlimited Solar.
No clouds. No night cycles (if placed in the right orbit).
Just pure, consistent energy — 8x more efficient than panels on Earth.
2. Free Cooling.
Space is a vacuum. No water consumption. No humidity.
None of the HVAC nightmares every Malaysian and Singaporean data center operator deals with.
3. No Land Constraints.
No zoning fights. No “please wait for the next masterplan.”
No need for 100 acres of land and a new substation.
Launch costs used to be the blocker, but not anymore:
2000s: $10,000/kg
Today: ~$1,500/kg
SpaceX target by 2035: ~$200/kg
And where your stomach drops:
At ~$200/kg, space-based data centers hit cost parity with certain AI workloads on Earth.
This is economics colliding with physics. (KaPow)
SEA: The Region With the Most to Gain and Lose
Let’s bring this home.
Right now, SEA is fighting an invisible war nobody wants to admit:
Malaysia: pausing new data center approvals because of electricity and water
Singapore: strict compute allocation because the grid is maxed
Indonesia & Vietnam: racing to plug infrastructure gaps “just in time”
Our region wants to be the AI hub of the East — but we’re trying to build a 2030 economy on 2010 infrastructure.
Google’s move is essentially saying:
“If countries can’t supply the power or cooling we need… we’ll go above them.”
That should be a wake-up call, but a window.
Because Southeast Asia has a unique strategic advantage: the equator.
Equatorial regions are ideal for orbital paths and ground station networks.
Whoever builds these first becomes the access point for orbital compute. One win, at a time.
This is the new “port city” moment, but instead of trade routes on water, it’s trade routes in space.
This Isn’t Just “Tech News” — It’s a Rerouting of Value
People hear “space tech” and imagine rockets, spacesuits, Elon in a leather jacket.
But the real impact is not in space.
It’s in the industries that feed families and employ millions here.
1. AI Compute: The Ceiling No One Talks About
Founders don’t publicly say this, but privately they admit:
GPU access is painful
Training runs are expensive
Data centers are oversubscribed
Even “cloud” is bottlenecked by physical infrastructure
Orbital compute changes this:
Cheap energy
No water for cooling
Stable environment
No land cost
Shorter upgrade cycles
Heavy training workloads can run in orbit.
Inference happens on Earth.
It’s the literally the best of both worlds.
2. Agriculture: SEA’s $400B Industry, Reinvented
This is where it gets beautifully human.
Picture a farmer in Chiang Mai, Long An, or Kelantan.
She plants based on gut instinct passed down from her parents — the weather, the soil, the “feel.”
Now imagine she gets:
Soil health insights down to micro-plots
Pest outbreak warnings 72 hours before they hit
Water stress signals before leaves even turn yellow
A planting window tailored to her exact microclimate
All powered by:
Satellite imaging
AI models running directly in orbit
Insights beamed down to her phone
And it’s not hypothetical. Startups like AcerX and LunaSat are already building toward it.
3. Mining, Utilities & Infrastructure: The Unsexy Big Winners
Mining companies can identify mineral deposits without sending teams into the field.
Utilities get real-time grid monitoring.
Supply chains gain route optimisation from orbit.
The boring industries the ones that actually move trillions just might benefit most.
and the BUT - The Hard Part: Space Is Unforgiving
Let’s not romanticise it.
Space is brutal. Very- too much space- no good.
Maintenance?
On Earth, data centers have technicians.
In space, unless Kevin’s secretly building a robotics lab, you need humanoid robots.
The funny thing is… they might actually arrive before half the “AI startups” raising money today.
Thermal Radiation?
Space is cold, but cooling is weird in a vacuum.
There’s no convection — only radiation.
So you need massive radiator panels that look more ISS than Instagram-aesthetic.
Latency & Bandwidth?
Yes, LEO can be faster than Singapore → California.
But you still need:
Laser interlinks
Dense ground stations
Spectrum coordination
Policies that don’t take 10 years to debate
This is where SEA has ~24–36 months to move.
After that, the infrastructure map locks in.
What Founders Should Actually Do (Without Quitting Everything to Start a Rocket Company)
Here’s the grounded version (time to be down-to- earth) HA! — actionable, human, sane:
1. Map your “Orbital-Eligible” Workloads
Which parts of your stack are:
compute-heavy
non-latency-critical
expensive to run
bottlenecked by infrastructure today
That’s your future orbital backlog.
2. Get to Know the Space-Tech Folks
You don’t need to build satellites, but you do need to understand what’s possible.
Grab coffee with teams like:
AcerX
LunaSat
One Orbit
StarCloud founders visiting the region
One coffee chat > 50 hours of reading PDFs.
3. Start Using Earth Observation Data Now
Don’t wait for “perfect.”
The companies that learn satellite data workflows today will dominate by 2030.
4. Add One Slide to Your 5-Year Strategy
Literally one slide:
“How does our model change if orbital compute becomes cost-competitive by 2029?”
This single question will separate resilient companies from fragile ones.
5. Join the Policy Conversations
Spectrum. Ground stations. Data flows.
This is no longer “for the government.”
It’s for anyone dependent on compute.
Be in the WhatsApp group, not the press release.
The Real Bet: Participate or Spectate
Every founder in SEA is making a bet — consciously or not.
If you ignore orbital compute, you’re betting that:
launch costs won’t drop
energy constraints won’t worsen
Big Tech won’t reshape the map
Historically, that’s not a great bet.
When cloud arrived, early adopters won.
When mobile arrived, early adopters won.
When space arrives quietly, then suddenly early adopters will win again.
We’re not saying “pivot to rockets.”
We’re saying “don’t get blindsided” go in with eyes wide open dude.
Southeast Asia has the talent, the geography, the hunger, and the emerging deep-tech ecosystem.
What we lack is urgency — and maybe a little imagination.
A Human Ending (And Kevin’s Hoverboard)
At one point in our recording, Kevin went on a passionate tangent about how he still hasn’t gotten the hoverboard Back to the Future promised him.
And honestly, that’s the perfect metaphor for this moment.
Every generation gets promised the future in a neat, shiny package.
Flying cars.
Robot butlers.
Teleportation.
Hoverboards.
Most of it doesn’t arrive on time.
Most of it arrives sideways, in shapes we didn’t expect.
Maybe the future wasn’t a hoverboard…
Maybe it was a data center quietly floating above us, catching sunlight no country can ration.
Maybe the future isn’t about “thinking big.”
Maybe it’s just about not spacing out when the big ideas finally show up.
And if Southeast Asia plays this right — participates, not spectates — we might not get Kevin his hoverboard…
…but we might just build the next great infrastructure platform of our time. Close enough lah.
And that brings us to you.
If you’re reading this and thinking “this doesn’t apply to me,” you’re probably wrong.
If you’re a startup founder:
Orbital compute will determine whether you can train your models, scale your workloads, or access affordable GPUs.
If you’re a corporate operator:
Your agriculture, logistics, energy, manufacturing, and supply chain decisions will increasingly rely on satellite-driven intelligence.
If you’re an investor:
The companies that understand this shift earliest — and position for it — will capture the compounding returns.
Everyone else will be late.
Everyone in Southeast Asia’s tech ecosystem is touched by this shift —
from the farmer in Vietnam using satellite insights,
to the AI engineer in Singapore fighting for GPU time,
to the logistics operator in Malaysia navigating floods, delays, and unpredictable demand.
This isn’t just a “Google thing.”
It’s a where-the-future-happens thing.
The question isn’t whether space-based compute is coming.
The question is:
Will Southeast Asia shape this future, or watch it happen from the sidelines?
— Kimberley & Kevin
What did we miss? What resonated?
Hit reply or drop a comment — these conversations shape how we cover the ecosystem.
🔗 Connect with us:
Kim: linkedin.com/in/weiisyuenyeoh
Kevin: linkedin.com/in/kevinbrockland
P.S.
Tags:
space tech, orbital compute, AI infrastructure, Google Project Suncatcher, Southeast Asia innovation, satellite analytics, agriculture tech, deep tech SEA, frontier tech, ground stations ASEAN, SpaceX, NVIDIA, climate tech, SEA startups, Elon Musk, Carl Sagan , Steve Jobs, Hoverboard, Backtothefutre
