The Pentagon has burned through 50,000+ rockets and missiles since Russia invaded Ukraine.
And suddenly… America can't make them fast enough.
So a strange thing is happening.
Silicon Valley startups aren't calling Lockheed for parts.
They're calling car factories.
And fracking rigs.
And pharma labs.
Meet Castelion — a California startup now valued near $3 billion.
Instead of buying aerospace-grade chips (slow, expensive, gatekept)…
It raided the auto industry.
Those Field-Programmable Gate Arrays that help your EV self-park?
Castelion uses them to guide hypersonic missiles.
Oh — and the high-pressure tubes inside those rocket motors?
They're borrowed from fracking rigs in Texas.
Same heat. Same pressure. Way more suppliers. Way lower price.
Then there's Anduril — the $61B defense darling.
Mixing rocket propellant used to look like a giant industrial KitchenAid. Paddles. Hours of stirring. Endless cleaning.
Anduril threw that out.
They bought a bladeless centrifugal mixer from Colorado's FlackTek — the same tech pharma companies use to make liposome cancer drugs.
👉 Multi-hundred-kilo propellant batches mixed in minutes, not hours.
👉 24x the output of conventional mixers.
👉 10x throughput vs. Anduril's own old rigs.
Because when you're mixing chemo drugs or missile fuel — one contamination ruins the batch.
X-Bow Systems in New Mexico is 3D printing solid rocket motors.
A new production line used to take 3–6 years to build.
X-Bow does it in 12 months. Pentagon already handed them $191M.
Firehawk Aerospace in Texas went even further:
🔥 Rocket fuel production dropped from 60 days → 7 hours.
At one-tenth the cost.
Even Northrop admits: swapping metal tooling for 3D-printed polymer cut factory setup from a year to six weeks.
The defense giants — Lockheed, Boeing, RTX — spent decades building bespoke, gold-plated supply chains.
The startups just asked a different question:
What if the tech we need already exists — in Detroit, in Odessa, in a pharma plant?
Washington is now dangling $53 billion and rewriting procurement rules to keep up.
The next arms race won't be won by the smartest engineer.
It'll be won by whoever remixes other industries the fastest.
And right now — the remixers are winning.
That's all for now!
