🏠

🌊 AI Waves: Floating Data Centers 🤯

May 06, 2026 | Author

Tech

🎧 Audio Summaries
English flag
French flag
German flag
Japanese flag
Korean flag
Mandarin flag
Spanish flag
🛒 Shop on Amazon

🧠Quick Intel


  • Panthalassa plans to deploy floating AI computing nodes in the Pacific Ocean by 2026, backed by investments from Silicon Valley investors like Peter Thiel totaling hundreds of millions of dollars.
  • The latest $140 million investment will support the construction of a pilot manufacturing facility near Portland, Oregon, and accelerate the deployment of wave-riding nodes.
  • Each Panthalassa node utilizes wave motion to drive a turbine generator, producing renewable energy to power onboard AI chips and cool the system with surrounding water.
  • The Ocean-3 prototype, scheduled for testing in the northern Pacific Ocean, will reach approximately 85 meters in length, comparable in height to London’s Big Ben.
  • Satellite links will facilitate data transmission, with potential bandwidths of hundreds of megabits per second per terminal, despite potential delays.
  • Panthalassa has previously tested prototypes including Ocean-1 (2021) and Ocean-2, which underwent a three-week sea trial off the coast of Washington state in February 2024.
  • Garth Sheldon-Coulson, CEO of Panthalassa, envisions deploying thousands of the nodes globally.

    • 📝Summary


    Panthalassa aims to deploy AI computing nodes in the Pacific Ocean, a venture backed by investors like Peter Thiel with hundreds of millions of dollars. The company intends to complete a manufacturing facility near Portland, Oregon, and accelerate the deployment of wave-riding nodes. These nodes, resembling large steel spheres, utilize wave motion to generate energy, powering onboard AI chips and transmitting data via satellite link. Earlier prototypes, including Ocean-1 tested in 2021 and Ocean-2’s February 2024 trial off Washington, are progressing. The latest version, Ocean-3, is slated for testing in the northern Pacific Ocean, potentially reaching a height comparable to London’s Big Ben. The concept transforms energy transmission into data transfer, relying on satellite communication for inference tokens, though bandwidth limitations remain a key challenge.

    💡Insights

    ▼


    CHAPTER 1: THE RISE OF OCEANIC AI
    Silicon Valley investors, including Peter Thiel’s Palantir, are injecting hundreds of millions of dollars into Panthalassa, a company pioneering floating AI data centers in the Pacific Ocean. This investment surge coincides with broader tech companies grappling with difficulties establishing traditional AI data centers on land, signaling a strategic shift towards novel computing solutions. The company’s recent $140 million funding round is intended to accelerate the manufacturing of pilot nodes and deployments leveraging wave energy.

    CHAPTER 2: PANTHALASSA’S INNOVATIVE NODE DESIGN
    Panthalassa’s core technology centers around large, steel sphere nodes designed to harness wave motion. These nodes utilize a vertical tube to draw water upwards, driving a turbine generator for renewable energy and simultaneously cooling the AI chips with the surrounding seawater. The Ocean-3 prototype, slated for 2026 testing in the northern Pacific, reaches approximately 85 meters in length – comparable in height to London’s Big Ben or New York City’s Flatiron Building. The company has already completed trials with earlier prototypes, including Ocean-1 (2021) and Ocean-2 (February 2024 sea trial off Washington state).

    CHAPTER 3: COOLING, COMMUNICATION, AND CHALLENGES
    The innovative cooling system, capitalizing on the ocean’s low ambient temperature, represents a significant advantage over conventional land-based data centers. However, the reliance on satellite links for data transmission introduces limitations regarding bandwidth and signal delays. The potential for coordination challenges between multiple nodes and the need for periodic data transport via ship pose further hurdles. Satellite communication is limited to around hundreds of megabits per second, while fiber-optic cables offer significantly faster data transfer.

    CHAPTER 4: A HISTORY OF UNDERWATER COMPUTING
    Panthalassa’s endeavor builds upon previous explorations in underwater computing. Microsoft’s Project Natick, which experimented with submerged servers in 2015 and 2018, demonstrated the viability of sealed, seawater-cooled systems with lower failure rates. Chinese companies have already deployed underwater data centers near Hainan Island and off Shanghai, and Keppel is constructing a floating data center for Singapore. These prior attempts, while ultimately unsuccessful in a commercial context, highlight the potential of leveraging the ocean’s environment for data processing.

    CHAPTER 5: SCALING AND THE FUTURE OF OCEANIC DATA
    Panthalassa’s CEO, Garth Sheldon-Coulson, envisions deploying thousands of nodes, but significant challenges remain. The autonomous and self-propelled nature of the nodes, initially deployed by ship, is crucial for maintenance and replacement. The company’s ambition to create nodes capable of surviving for over a decade without human intervention underscores the complexity of operating in the harshest ocean conditions. Despite the obstacles, the $210 million investment reflects a belief in the feasibility of this ambitious vision, particularly in comparison to alternative AI data center strategies.

    Related Articles