The Free News Press
Crosswords Sudoku and Comics
Business

Startup Raises $210 Million to Build AI Data Centers That Float on Ocean Waves

Panthalassa, backed by Palantir co-founder Peter Thiel, plans to deploy its first pilot nodes in the northern Pacific Ocean later in 2026.

Startup Raises $210 Million to Build AI Data Centers That Float on Ocean Waves
Startup Raises $210 Million to Build AI Data Cent…      Panthalassa Floating Data Center    Pixabay (free for editorial use)
By Free News Press Editorial Team
Published May 15, 2026 at 1:58 PM PDT

A Silicon Valley startup is betting that the next frontier for artificial intelligence computing is not a desert campus or a suburban office park but the open ocean. Panthalassa has raised $140 million in a Series B funding round to build autonomous, floating AI computing nodes powered by ocean waves, bringing its total funding to $210 million.

The round was led by Peter Thiel, the co-founder of Palantir, according to Fox News. The company says it will use the money to complete a pilot manufacturing facility near Portland, Oregon, and to deploy its Ocean-3 pilot node series in the northern Pacific Ocean later this year.

The basic idea is to move AI computing offshore, away from the land-based constraints that have made building new data centers increasingly difficult. On land, AI facilities require vast amounts of electricity, large plots of land, industrial cooling systems, and the cooperation of local communities that frequently oppose construction. Ocean-based nodes, Panthalassa argues, solve several of those problems at once.

Each floating node is designed to capture wave motion and convert it into electricity. Seawater handles cooling. Onboard chips process AI tasks and send results back to land through low-Earth-orbit satellites. The company says it spent a decade developing the underlying technology for power generation, onboard computing, and autonomous ocean operations. Earlier prototypes called Ocean-1, Ocean-2, and Wavehopper were tested in 2021 and 2024.

The specific AI workload Panthalassa is targeting is called inference. That is the step in AI where a trained model responds to a user's prompt, such as when a person types a question into a chatbot and receives an answer. Inference is generally less demanding than training a model from scratch, which requires massive, tightly coordinated data movement. Running inference at sea is considered more realistic than attempting to train large models on a floating platform.

The scale of electricity demand from AI has pushed companies and investors toward unconventional solutions. Data centers supporting large AI models consume power at a rate that strains regional electrical grids, and the demand is growing faster than new land-based generation capacity can be built. Ocean wave energy, while still an emerging technology, offers a source of electricity that does not compete with residential or industrial users on land.

Panthalassa has been developing its technology largely out of public view. The Series B round signals that institutional investors are willing to commit serious capital to the concept. Thiel's involvement adds a prominent name to what remains an unproven approach at commercial scale. The company's next visible test will come when the Ocean-3 nodes are deployed in the Pacific later in 2026.

Panthalassa Floating Data Center    Pixabay (free for editorial use)