Hundreds of miles from the nearest power line, a steel hull the size of a barge rises and falls with the swell. Inside, turbines spin, generators hum, and a rack of AI chips quietly answers questions sent from people who will never know where the answer came from. There is no grid here. There is no cable to shore. There is only weather, and Panthalassa has decided weather is the point.
This is the company as it stands in 2026: a 75-person public-benefit corporation in Portland, Oregon, fresh off a $140 million Series B, building floating machines that generate electricity in the open ocean and spend it on the spot. They do not ship the power home. They ship the homework.
"There are three sources of energy on the planet with tens of terawatts of new capacity potential: solar, nuclear, and the open ocean."
- Garth Sheldon-Coulson, Co-Founder & CEOThe math is unglamorous. The world wants more compute than it can plug in. Data centers queue for years to connect to transmission lines that were never built for this kind of appetite, in regions where the public would rather not host another humming warehouse. Energy is abundant, but it is rarely abundant where you happen to need it.
Most companies respond to this by fighting over the same crowded coastline of land, permits, and substations. Panthalassa looked at the same map and noticed that two-thirds of it is water - and that the windiest, wave-battered stretches of ocean, the parts nobody wants to be near, are precisely where the energy is densest. Inconvenient, yes. Unoccupied, also yes.
"The future demands more compute than we can imagine. Panthalassa opened the ocean frontier."
- Peter Thiel, lead investor, Series BGarth Sheldon-Coulson and co-founder Brian Moffat started Panthalassa in 2016 with a contrarian inversion. For a century, energy infrastructure has been built around one assumption: generate power somewhere, then transmit it to where people are. The whole apparatus of grids, cables, and substations exists to close that distance.
Their bet was to delete the distance entirely. If the most valuable thing you can do with electricity is run AI inference - and inference is just math that travels at the speed of light - then you do not need wires at all. You need a satellite uplink. Put the generator and the computer in the same floating hull, let the ocean cool the chips for free, and send the answers home as data instead of electrons.
It took nearly a decade of building propulsion, autonomy, and energy systems to make that sentence stop sounding like science fiction. The team they assembled - veterans of SpaceX, NASA, Apple, Boeing, Tesla and Blue Origin - treats a power plant less like a utility asset and more like a spacecraft that happens to float.
"Renewably generating the lowest-cost electrons on Earth - and never bothering to bring them ashore."
- The Panthalassa premise, paraphrasedPanthalassa builds autonomous nodes - floating energy systems mass-produced from plate steel in coastal factories, the way a shipyard turns out hulls rather than the way a utility commissions a one-off plant. Each node drifts in an energy-dense ocean region and harvests the motion of the waves: as the structure rises and falls, fluid flows through internal turbines, spinning generators that make electricity around the clock.
That electricity never sees a transmission line. It powers AI inference chips mounted inside the node, with the surrounding seawater doing the unglamorous job of supercooling that data centers on land pay fortunes to engineer. The results - inference tokens - travel back to land by low-Earth-orbit satellite. The ocean does the hard part; physics handles delivery.
"Go where the energy is."
- Company taglineSkepticism is the correct first reaction. The ocean destroys things; that is its hobby. So the argument has to rest on numbers, and the central number is the capacity factor - the share of the time a power source is actually producing. This is where the open ocean stops looking eccentric and starts looking obvious.
Then there is the money, which tends to settle arguments faster than physics. In May 2026 Panthalassa closed a $140 million Series B led by Peter Thiel, joined by an investor list that reads like a who's-who of people who like being early: Founders Fund, John Doerr, Marc Benioff's TIME Ventures, Max Levchin's SciFi Ventures, Figma's Dylan Field, Gigascale Capital, Lowercarbon Capital, Super Micro Computer, and more. The round funds the pilot manufacturing facility near Portland and the Ocean-3 deployment.
"Panthalassa's autonomous wave power system addresses global energy needs - and strengthens American technological leadership."
- John Doerr, investorPanthalassa registered as a public-benefit corporation, which is a polite way of saying the mission is written into the paperwork. The stated goal is a planetary-scale energy platform that is clean, sustainable, and low-cost - one that meets surging demand without asking a single new community to host it, and without adding a gram of emissions.
The deeper claim is about scale. Solar and nuclear can grow enormously, but the open ocean is the third terawatt-class resource that almost nobody is industrializing. If even a sliver of it can be turned into reliable, mass-produced compute, the ceiling on clean energy moves somewhere most people have never thought to look.
The hard questions are still open. The ocean is brutal, manufacturing at fleet scale is unproven, and "we'll deploy commercially in 2027" is a sentence every hardware startup has said at least once. Panthalassa has prototypes and a marquee cap table, not yet a running business. Skeptics are entitled to keep their receipts.
But the direction is worth taking seriously. If the bet works, the lonely stretch of the North Pacific stops being a void on the map and becomes infrastructure - a place that makes electricity, thinks, and sends its thinking home, without a wire, a permit fight, or a neighbor to object. The same swell that ships have dreaded for centuries becomes the reason the lights stay on for an industry that never sleeps.
Go back to that steel hull rising and falling in the swell, hundreds of miles from anywhere. A year ago it was an empty, miserable patch of ocean. Now it is answering questions. Panthalassa's wager is that there are tens of terawatts more where that came from - and that the rest of us simply forgot to look offshore.
"The ocean never clocks out. Panthalassa decided that was a feature, not a hazard."
- YesPress