Panthalassa raises $140M to build autonomous ocean energy nodes for off-grid AI inference

Panthalassa raises $140M to build ocean energy nodes for off-grid AI inference

Panthalassa has closed a $140 million round to scale manufacturing of autonomous ocean energy nodes that capture wave and wind energy in the Southern Hemisphere and use it to run AI inference workloads off-grid.

Founder Garth Sheldon-Coulson describes each node as generating roughly 500 kilowatts, with a range of 100 kilowatts up to a megawatt — enough to power one rack up to several, depending on density. The nodes communicate via radio and satellite rather than fiber, which rules them out for synchronous model training but makes them well-suited for large-scale parallel inference. Sheldon-Coulson says the company is already being asked to plan toward 100 GW per year of energy capacity, framing the long-term build as a pure manufacturing scale problem.

“The typical node will be on the order of 500 kilowatts... they can talk to each other sideways. So we'll have radio. Of course, they're talking to satellite... If you want tons of embarrassingly parallel inference, running all the same models, running future models that get bigger or smaller, we think we're going to be perfect for lots of that... 10 GW, 100 GW per year is the kind of thing that we're getting asked to do.”

Why the Southern Hemisphere

The target deployment zone is the Southern Hemisphere ocean belt — south of Australia and New Zealand, circling the planet well north of Antarctica. Sheldon-Coulson argues this resource is unusually attractive because the wind blows almost constantly and the ocean acts as a battery for both wind and solar energy, keeping nodes online close to four nines of uptime in optimized configurations. Battery storage is on-board but rarely cycled — unlike solar, which draws on storage daily, these nodes dip below full generation only a few times per year.

Manufacturing path

Panthalassa's first pilot production line is running near Portland. The plan is to dial in manufacturing there, then replicate those lines in countries near the Southern Hemisphere deployment zones to reduce logistics costs. Sheldon-Coulson says the company is already in conversations with partners in relevant countries to identify factory sites.

What unlocked the round

Two things converged. Panthalassa has now developed the core technology to a point where it can begin scaling — no autonomous ocean energy capture system existed before — and the broader energy constraint for AI has become impossible to ignore. Data center power is scarce, GE Vernova's grid equipment is backordered, and natural gas remains the default alternative. The $140 million goes toward standing up manufacturing capacity and deploying the first commercial fleet later this year.