Compute Follows the Power

A Peter Thiel backed venture raised capital to run AI compute on wave powered platforms at sea, showing how far infrastructure will travel to reach available energy.

What Happened

Panthalassa, an ocean technology company based in the United States, has raised fresh capital that places its valuation near one billion dollars. Peter Thiel led a funding round of 140 million dollars, joining a venture that has spent roughly a decade developing wave energy systems. The company now says it is ready to build factories and deploy fleets at scale.

The plan places both power generation and computation on the open ocean. Floating steel structures, which the company calls nodes, use the motion of waves to drive water through a turbine and generate electricity directly. Each node measures around 85 metres in length and sits mostly below the surface, with a sealed server compartment cooled by the surrounding seawater.

These platforms are designed to run without any connection to shore. Rather than sending electricity back to land, they receive user queries through satellite links and return results the same way. Pilot units are planned for the northern Pacific this year, with wider commercial deployment targeted for 2027.

Structural Context

The venture reflects a broader shift in how the compute industry thinks about growth. Demand for AI processing continues to accelerate, while the systems that deliver power to large sites face real limits. Across many markets, the binding constraint on new capacity is no longer capital or chips. It is access to electricity and the transmission needed to move it.

This is why so much recent activity points outward, toward the source of energy rather than the traditional map of grid connection points. Interconnection queues stretch across years in several regions, and permitting timelines for large loads remain long. When power becomes the scarce input, infrastructure begins to travel toward wherever generation can be created or captured. Placing compute at sea is one striking expression of that logic.

The ocean carries enormous theoretical energy, and the appeal is easy to understand. At the same time, the marine environment introduces demanding conditions. Saltwater corrosion, biofouling, storm exposure, and the logistics of servicing remote hardware are all genuine engineering questions that any offshore programme will need to resolve as it scales.

The Enki Perspective

Panthalassa validates a principle that sits at the centre of how Project Enki reads the market. Compute follows power. When the grid cannot deliver energy quickly enough, the most capable operators stop waiting for it and start building where energy already exists. The direction of travel is toward generation, not away from it.

Project Enki applies the same underlying logic to a different and immediately accessible resource. Across Europe, renewable generation is frequently curtailed or stranded because transmission cannot absorb it. That surplus represents real economic value that currently goes unused. By siting modular compute directly at these locations, Enki turns energy that would otherwise be wasted into scalable AI capacity, without waiting for new transmission to be built.

Offshore programmes and onshore stranded energy strategies are complementary parts of the same story. Both recognise that power availability now defines where digital infrastructure can grow. Enki's focus on curtailed renewables inside European jurisdictions adds a further dimension, because infrastructure anchored to domestic generation strengthens regional control over compute. Digital sovereignty, in this reading, rests on energy sovereignty.

What This Signals

The scale of capital moving into offshore compute confirms that the energy constraint is now a defining feature of the AI build out, not a temporary friction. Serious investors are funding ambitious approaches precisely because demand for compute is expected to keep climbing well beyond what conventional grid expansion can support.

For institutional capital, the lesson is about where durable value will be created. Repeatable infrastructure models that pair compute with reliable generation are likely to attract sustained financing, because they address the actual bottleneck. Whether power comes from waves far offshore or from curtailed renewables closer to demand, the leading designs will be those that shorten the path between electricity and computation.

The broader signal is that the geography of compute is being redrawn around energy. As that shift continues, value will accrue to infrastructure that sits closest to available power and can be deployed at speed. Project Enki is positioned squarely within that evolution.

Source: Efosa Udinmwen,
7 May 2026 https://www.techradar.com/pro/were-now-ready-peter-thiel-backed-company-raises-usd1bn-to-send-data-centers-out-to-sea-to-harness-tens-of-terawatts-of-new-capacity-potential-in-the-power-of-the-open-ocean