What is an ocean data center? An ocean data center is a floating compute facility deployed on the open sea that draws power from wave energy and uses cold seawater for cooling, eliminating the land, grid, and water constraints that increasingly bottleneck terrestrial AI infrastructure. On May 4, 2026, Oregon-based Panthalassa announced a $140 million Series B led by Peter Thiel to scale its wave-powered ocean data center technology, joined by John Doerr, Marc Benioff's TIME Ventures, Founders Fund, Lowercarbon Capital, Super Micro Computer, and Hanwha Asset Management. For CRE investors who have spent the last 24 months chasing land for hyperscale data center campuses, the announcement signals a parallel asset class taking shape offshore. For a comprehensive map of how AI infrastructure decisions are reshaping CRE, see our complete guide to AI commercial real estate.
Key Takeaways
- Panthalassa raised $140 million in Series B funding led by Peter Thiel to deploy wave-powered ocean data centers, pushing the company close to a $1 billion valuation.
- Each node is a buoyant lollipop-shaped structure that converts wave motion into electricity and uses cold seawater for cooling, sidestepping land and grid constraints.
- Pilot manufacturing near Portland, Oregon will support deployment of Panthalassa's Ocean-3 nodes in the northern Pacific in 2026, with commercial deployments targeted for 2027.
- The funding signals that LP capital is willing to back alternative data center asset classes as land-based hyperscale projects face permitting and community pushback.
- For CRE investors, ocean data centers are unlikely to displace terrestrial campuses but will compete for the same AI inference workloads, especially latency-tolerant ones.
Ocean Data Center CRE Investors Need to Understand the Model
Panthalassa's technology pairs wave power generated by floating orbs with onsite AI computing, transmitting data via low-Earth-orbit satellites. Each lollipop-shaped node consists of a buoyant spherical head connected to a long submerged vertical tube. As ocean waves pass, the node bobs up and down, while the surrounding water moves in small orbital paths. That relative motion drives oscillating flow through internal turbines, which generate electricity onsite to power servers and cool them with cold seawater. The architecture eliminates the costly transmission lines that normally connect ocean energy to land-based loads.
CEO Garth Sheldon-Coulson framed the bet this way: there are three sources of energy on the planet with tens of terawatts of new capacity potential, namely solar, nuclear, and the open ocean, and Panthalassa is building for the third. Peter Thiel called the ocean a frontier for compute that is no longer science fiction. The company has been operating as a public benefit corporation since 2016, with prior prototypes Ocean-1, Ocean-2, and Wavehopper deployed in sea trials during 2021 and 2024.
Why Ocean Data Centers Are Emerging Now
Three forces converged in 2026 to make ocean data centers viable financing targets for venture capital and, eventually, real asset investors.
First, land-based data center sites are saturated in tier-one markets. Northern Virginia, Phoenix, Dallas, and the Hillsboro corridor are all running into permitting friction, transmission queue delays, and rising community opposition. The Redfin survey on AI data center neighborhood pushback found that 47% of Americans oppose AI data centers in their communities. Power interconnection queues stretch 4 to 7 years in many regions, while a wave-powered ocean node has no grid dependency at all.
Second, hyperscaler capex is exploding. Meta's Q1 2026 earnings disclosed $145 billion in 2026 AI capex, and Amazon committed $44 billion in Q1 alone. With this much capital chasing compute, alternative siting strategies become economically rational even at higher per-megawatt build costs. For deeper context on hyperscaler spending, see our analysis of Meta's $145B 2026 capex plan.
Third, water and cooling constraints are biting. A typical hyperscale facility consumes millions of gallons of water annually for evaporative cooling, and several jurisdictions including Texas and Arizona are tightening industrial water use rules. Panthalassa's submerged architecture uses ambient seawater, eliminating freshwater draw entirely.
Implications for CRE Data Center Investors
Ocean data centers will not replace land-based campuses for the workloads that demand sub-5 millisecond latency to dense urban populations, like real-time inference for ad targeting or trading. But for AI inference workloads where 50 to 150 millisecond latency is acceptable, including batch model evaluation, video processing, and most agentic AI workflows, the offshore model competes directly. CRE investors should think about exposure across three dimensions.
Land valuations may peak in saturated markets. If even 10% of incremental AI inference capacity moves offshore by 2030, the bid for land in tier-one markets weakens. Investors holding optioned land or pursuing entitlements in oversubscribed corridors should stress-test deals against a slower land absorption pace. The dynamics here mirror what we saw with Coatue's $5.7B Next Frontier land venture, which signals continued private market appetite for land but increasingly disciplined pricing.
Coastal industrial real estate gains optionality. Ocean data centers need shore-side support: pilot manufacturing facilities like Panthalassa's Portland operation, fiber landing stations, vessel maintenance yards, and crew transfer infrastructure. Coastal industrial parcels in Oregon, Washington, Northern California, the Gulf Coast, and the New England coastline could see new tenant demand from offshore compute companies. CBRE Research tracks the industrial coastal supply chain that would underpin this.
REIT exposure shifts. Pure-play data center REITs like Equinix and Digital Realty are not threatened by 2027 deployments, but allocators thinking 5 to 10 years out should track which platforms acquire offshore capabilities. The same underlying AI demand that drives terrestrial REIT capex also drives Panthalassa's roadmap. For investors building a balanced AI infrastructure portfolio, our guide to AI tools for real estate investors covers how to think about platform diversification.
Risks and Open Questions
Ocean data centers face execution risk that terrestrial CRE underwriters are not used to pricing. Maintenance costs at sea, biofouling on submerged structures, vessel availability for service runs, and salt-water corrosion are real engineering challenges. Regulatory approval for offshore compute siting in U.S. waters is also untested, with overlapping jurisdictions across the Bureau of Ocean Energy Management, NOAA, and state coastal commissions. Insurance markets for floating compute infrastructure are also still developing.
For personalized guidance on positioning your CRE portfolio against the rise of alternative data center asset classes, connect with The AI Consulting Network. CRE investors looking for hands-on AI implementation support can reach out to Avi Hacker, J.D. at The AI Consulting Network for a customized infrastructure strategy review.
Real-World Applications and Market Context
The AI in real estate market is forecast to reach $1.3 trillion by 2030, growing at a 33.9% CAGR. Within that, data centers represent the single largest infrastructure line item, with U.S. data center spending estimates exceeding $250 billion in 2026 driven by hyperscaler capex commitments. CRE sales volume is forecast to increase 15 to 20% in 2026, and a meaningful share of that activity is data-center-adjacent. Panthalassa is one of several alternative data center plays raising serious capital. Combined with terrestrial activity like the Fleet Data Centers $4.6B Nevada campus and the Meta El Paso 1GW campus, the message for CRE investors is clear: AI infrastructure is becoming a multi-modal asset class that includes land, sea, and eventually low-earth orbit. If you are ready to transform your AI infrastructure thesis, The AI Consulting Network specializes in exactly this work.
Frequently Asked Questions
Q: How does an ocean data center compare to a land-based hyperscale campus?
A: Ocean data centers eliminate land, grid interconnection, and freshwater cooling constraints, but they introduce maintenance complexity, latency penalties for distant users, and untested regulatory pathways. They are best suited for latency-tolerant AI inference workloads rather than real-time consumer applications.
Q: When will Panthalassa's ocean data centers be commercially operational?
A: Panthalassa is targeting deployment of its Ocean-3 series of nodes in the northern Pacific Ocean in 2026, with commercial deployments planned for 2027. The current $140 million round funds completion of its pilot manufacturing facility near Portland, Oregon.
Q: Should CRE investors revalue land-based data center exposure because of this?
A: Not in 2026. The current ocean data center capacity is too small to materially affect land bids in tier-one markets. However, allocators thinking 5 to 10 years out should treat offshore compute as a real demand-substitution risk for the most saturated and least-permittable land markets.
Q: What CRE asset classes benefit if ocean data centers scale?
A: Coastal industrial real estate near pilot manufacturing hubs, fiber landing stations, port-adjacent service yards, and crew accommodation properties all gain optionality. Oregon, Washington, the Gulf Coast, and parts of the New England coastline are early candidates for shore-side support tenant demand.
Q: Is Panthalassa the only ocean data center company?
A: Panthalassa is the largest funded entrant as of May 2026, but the broader category includes earlier projects like Microsoft's Project Natick prototype and several stealth-mode startups exploring submerged compute. The Panthalassa round signals that institutional capital is now willing to underwrite the category at meaningful scale.