PayPal co-founder Peter Thiel has placed a $140 million bet on ocean-based infrastructure and has committed to a start-up building steel nodes the length of a football pitch that float in open ocean and power AI servers using wave energy. The new funding from the billionaire investor will allow the start-up to finish building its pilot manufacturing facility near Portland, with Panthalassa developing technology that pairs wave power generated by massive floating structures with on-site AI computing. The systems then transmit results to shore via low-Earth-orbit satellites, cutting the need for costly undersea cables entirely.

What sets Panthalassa apart from other floating data centre projects is its energy model. The company does not draw power from shore or connect its nodes to undersea cables. Instead, the nodes generate clean electricity directly from wave motion, then use that power onboard to run AI chips, with the surrounding ocean also serving as a natural cooling system, which Panthalassa describes as free supercooling.

Keeping chips cool is one of the most expensive operational challenges facing conventional data centres. Panthalassa’s approach addresses it without a single cooling tower, chiller unit, or drop of municipal water.

The land problem that nobody wants to talk about

The funding did not emerge from a vacuum. Behind the headlines is a data centre industry under acute strain. Terrestrial data centres face mounting constraints: limited grid capacity, cooling water scarcity, supply chain bottlenecks, permitting delays, and impacts on local communities and infrastructure.

In parts of the United States, Ireland, and the Netherlands, grid operators have effectively stopped accepting new connection requests from large compute facilities. The time between site selection and power-on for a greenfield data centre in many Western markets now routinely exceeds three years.

AI data centre planned capacity in the United States alone now exceeds 50 gigawatts of demand, a figure that is straining electricity networks not designed for such concentrated load. Wave energy, by contrast, remains almost entirely untapped.

Wave energy along the US coast alone is estimated at 2,500 terawatt-hours per year, a figure that dwarfs current projections for offshore wind in the same geography. Garth Sheldon-Coulson, Panthalassa’s co-founder and chief executive, has been making this case for years. “The ocean is really unlimited in terms of how much energy is available,” he told CBS. “It will really be the cheapest energy on the planet.” That is a bold claim, but one that is becoming harder to dismiss as grid constraints tighten on land.

Sheldon-Coulson was equally direct in Monday’s announcement. “We’ve built a technology platform that operates in the planet’s most energy-dense wave regions, far from shore, and turns that resource into reliable clean power,” he said. “We’re now ready to build factories, deploy fleets, and provide a sustainable new source of energy for humanity.”

Thiel’s own words were characteristically expansive. “The future demands more compute than we can imagine,” he said. “Extra-terrestrial solutions are no longer science fiction. Panthalassa has opened the ocean frontier.”

How the nodes actually work

Panthalassa’s floating units, which it calls “nodes,” are large steel structures about 85 metres long. Most of each unit sits below the surface and includes a sealed container holding computer servers, cooled by seawater. The Ocean-3 functions similarly to a floating hydroelectric dam, with water cycled into a ballast tank as the platform rises and falls with the waves, forcing it through an internal turbine to generate electricity.

Critically, the units are self-propelled. The nodes themselves can use the shape of their hull to navigate autonomously, powered by ocean waves, to targeted deep-sea areas – eliminating the grid connection bottleneck of land-based data centres. The company asserts that, if scaled, its platform could provide power at costs potentially as low as $0.02 per kWh, a figure that would undercut virtually every land-based power purchase agreement currently being signed in the hyperscale market.

The investor syndicate behind the round reflects how seriously the broader tech and energy establishment is taking this. New participants include John Doerr, Marc Benioff’s TIME Ventures, Max Levchin’s SciFi Ventures, Susquehanna Sustainable Investments, Hanwha Group, Super Micro Computer, and Planetary VC, among others. Returning investors include Founders Fund, Gigascale Capital, and Lowercarbon Capital.

Engineering realities and what comes next

The sceptics are not without ammunition. Challenges remain, including potential saltwater corrosion, biofouling, and operational reliability during storm conditions. Satellite connections may also introduce latency issues, and repairing hardware in open ocean is a different proposition from dispatching a technician to a campus facility.

These are genuine concerns, and the company has been careful not to oversell what the Ocean-3 deployment represents. The capital will complete Panthalassa’s pilot manufacturing facility near Portland, Oregon, and fund initial deployment of its Ocean-3 node series in the northern Pacific in 2026, ahead of targeted commercial roll-out in 2027.

For those watching from the shore, the more interesting question may not be whether Panthalassa succeeds on its current timeline; it is whether the fundamental idea of moving compute off the grid and onto the water has now crossed a credibility threshold from which it will not retreat.

Thiel has been wrong before, and he has been spectacularly right before. What is less in doubt is that the land-based data centre industry’s core assumptions – cheap land, available power, cooperative planning authorities – are under pressure that is only going to intensify as AI workloads continue to scale. Panthalassa is betting that the ocean is the answer. With $140 million and some of Silicon Valley’s most recognisable names behind it, that bet has just become considerably harder to ignore.

RELATED STORIES

AI data centres could triple electricity consumption by 2030, IEA warns