Two college juniors built a $270M revenue data center equipment company on $3.4M while everyone else is raising billions.
The Summary
- Giga Energy, founded by two Texas A&M students in 2019, has generated over $270M in revenue while raising just $3.4M in equity and under $5M in debt
- The company manufactures critical power infrastructure (transformers, switchgear) for AI data centers, supplying clients like CoreWeave
- Giga now plans to build its own Nvidia-based AI data centers, which will likely force them to tap capital markets they've avoided until now
The Signal
While Crusoe Energy raises $3 billion and Applied Digital burns through funding rounds, Giga Energy built a quarter-billion-dollar revenue business on college-kid capital. The difference isn't luck. It's where they started in the stack.
Giga sells picks and shovels: transformers and switchgear that regulate power flow. These are unsexy, capital-light products with actual margins. You manufacture, you sell, you collect. No need to lease land, build facilities, or negotiate power purchase agreements that tie up capital for years. The company essentially bootstrapped by solving the boring infrastructure problem everyone else outsources.
But here's the tension: Giga is now moving downstream into building its own Nvidia-based data centers. That's the capital-intensive game they've avoided. CEO Matt Lohstroh acknowledges they'll need to raise soon. The question is whether their manufacturing cash flow gives them leverage other data center startups don't have. If you control your own power infrastructure supply chain, your cost basis is fundamentally different. You're not paying another company's margin on every transformer.
This matters because AI infrastructure is bifurcating. One path: raise billions, build big, hope the hyperscalers sign long-term contracts. The other: control a profitable piece of the value chain first, then expand from strength. Giga took the second path. Whether that translates when they start competing for AI workloads against well-funded competitors is the test ahead.
The Implication
Watch whether Giga's capital raise, when it comes, is materially smaller than peers because of their equipment business cash flow. If they can build data centers at lower cost because they manufacture their own infrastructure, that's a real moat. If not, they're just another data center startup that happened to delay the inevitable capital raise. The difference between those two outcomes will tell you whether vertical integration still matters in the age of infinite AI capital.
Source: The Information