admin 
Panthalassa has secured $140 million in Series B financing to push artificial intelligence computing into open waters, backing a plan to run inference workloads on floating platforms powered by ocean waves rather than land-based grids.
Peter Thiel led the round, joined by John Doerr, Marc Benioff’s TIME Ventures, Max Levchin’s SciFi Ventures, Susquehanna Sustainable Investments, Hanwha Asset Management’s venture fund, Fortescue Ventures, Future Positive, Super Micro Computer and other investors. Existing backers Founders Fund, Gigascale Capital, Lowercarbon Capital, Unless and WovenEarth also participated, taking total capital raised by the company to about $210 million.
Portland-headquartered Panthalassa plans to use the fresh funding to complete a pilot manufacturing facility near Portland and accelerate deployment of its Ocean-3 nodes, autonomous floating systems designed to generate electricity from wave motion and use it directly onboard to power AI chips. The company says the first Ocean-3 pilot series is planned for the northern Pacific Ocean in 2026, with commercial deployments targeted for 2027.
The proposition is aimed at one of the most urgent constraints facing the AI industry: power. Data centre electricity consumption is projected to more than double to about 945 terawatt-hours by 2030, with AI a central driver of that growth. That level would be slightly above Japan’s current electricity use, while data centre power consumption is expected to grow around 15 per cent a year between 2024 and 2030.
Panthalassa’s model seeks to avoid some of the most difficult bottlenecks confronting large AI data centres, including grid interconnection delays, limited power availability, land constraints, cooling-water pressure and local resistance to huge energy-consuming facilities. Instead of transmitting power back to shore, each floating node would generate electricity at sea, run computing equipment onboard and send inference output back to land through low-Earth-orbit satellite links.
Garth Sheldon-Coulson, the company’s co-founder and chief executive, said the open ocean is one of the few energy sources with “tens of terawatts” of potential new capacity. The company has described its platform as operating in high-energy wave regions far from shore, where autonomous systems can move, generate power and perform computing without cables or anchors.
The technical design resembles a floating hydroelectric system. As waves lift and lower the platform, internal water movement drives turbines that generate electricity. The surrounding ocean also offers cooling advantages, a key issue for AI infrastructure where advanced chips produce intense heat and require stable thermal management. Panthalassa says it has spent about a decade developing power generation, propulsion, autonomy and at-sea computing capabilities, with Ocean-1, Ocean-2 and Wavehopper prototypes tested in 2021 and 2024.
Thiel framed the investment as a bet on a new frontier for computing infrastructure, saying future compute demand will exceed present expectations. Doerr described autonomous wave power as a strategic clean-energy asset. Their involvement gives Panthalassa financial credibility at a moment when investors are searching for ways to connect the AI buildout with firm, low-carbon electricity.
The company’s broader investor group also reflects the convergence of climate technology, venture capital and AI hardware. Super Micro Computer’s participation is notable because offshore computing will require ruggedised systems capable of operating in harsh marine conditions. Hanwha, Fortescue and other industrial investors point to interest beyond software, including energy systems, manufacturing and maritime infrastructure.
However, the concept faces major execution risks. Saltwater corrosion, biofouling, storms, mechanical fatigue and maintenance costs have challenged wave-energy developers for decades. Offshore repair work is expensive, weather-dependent and operationally complex. Satellite connectivity could also make the platform better suited to inference tasks than latency-sensitive training workloads, limiting the range of AI jobs that can be handled at sea.
Regulatory and environmental scrutiny is also likely to grow if deployments move from pilots to fleets. Large numbers of autonomous offshore computing nodes would raise questions over navigation, marine habitats, fisheries, emissions accounting, data security and maritime jurisdiction. Panthalassa’s advantage will depend not only on whether it can generate low-cost power, but also on whether it can operate reliably at scale without creating new risks for ocean ecosystems.
Competition is widening as the AI sector searches for alternatives to conventional data centres. Offshore wind-linked data centres, underwater facilities, nuclear-powered campuses, gas-backed AI parks and even space-based solar power proposals are moving from concept papers to investment pipelines. Panthalassa’s pitch stands out because it places both power generation and computing at the energy source, removing the need for grid delivery.
