NGE · Investment Letter · Issue 05 · June 2026

The Ocean Compute.
Panthalassa, Peter Thiel,
and the data centre
that floats.

In May 2026, Peter Thiel led a $140 million round into a company building wave-powered floating data centres in the Pacific Ocean. John Doerr and Marc Benioff followed. When these three invest together, it is worth paying attention. This letter explains what Panthalassa is, why it matters, and what it means for the $1Q thesis.

Not investment advice. Not a recommendation to buy or sell. Research and long-horizon thinking only. Consult a qualified financial advisor before making any investment decision.

The Signal

When Thiel leads a round,
the technology works.

Peter Thiel backed Facebook at $500,000 when it had one million users. He backed Palantir before anyone in the mainstream understood what it was. He backed SpaceX when rockets were still exploding. His investment thesis is not consensus — it is conviction about things that look improbable but are technically real.

In May 2026, Thiel led a $140 million Series B into Panthalassa — an Oregon-based startup building floating data centres powered entirely by ocean waves, deployed in the northern Pacific Ocean. John Doerr, the venture capitalist who backed Google and Amazon, co-invested. Marc Benioff's personal fund joined. The valuation is approaching $1 billion.

These are not tourists. These are people who have seen the hardware in the water and decided it works. When Thiel said at the announcement — "Extra-terrestrial solutions are no longer science fiction. Panthalassa has opened the ocean frontier" — he was not doing marketing. He was stating a technical conclusion.

$140M
Series B led by Peter Thiel · May 2026
~$1B
Valuation · approaching unicorn status
2027
Commercial deployment target · Pacific Ocean
What It Is

A data centre that runs
on the Pacific Ocean itself.

Panthalassa builds autonomous floating platforms — the Ocean-3 series — that do two things simultaneously. They generate electricity from ocean waves using a proprietary wave energy conversion system. And they use that electricity to run AI inference computing onboard, transmitting data via low-Earth-orbit satellites.

The ocean provides the power. The ocean provides the cooling — seawater is a far more efficient heat sink than any air-cooling system on land. The platform is autonomous — no crew, no shore connection, no grid dependency. And it deploys in international waters, which means no permitting delays, no community opposition, no land acquisition costs.

01

Wave Energy → Electricity

Massive floating orbs convert ocean wave motion into electricity. The northern Pacific has some of the most energy-dense wave regions on earth — consistent, powerful, and entirely unoccupied. No fuel. No grid. No emissions.

02

Seawater Cooling

The single largest operating cost of a terrestrial data centre is cooling. Panthalassa's platforms use seawater directly — essentially infinite cooling capacity at near-zero marginal cost. This alone changes the unit economics of AI compute fundamentally.

03

AI Inference at Sea

The platforms run AI inference computing onboard — the process of running AI models to generate outputs. This is the most electricity-intensive part of the AI value chain, and the part growing fastest. Move the inference to where the power is cheapest.

04

LEO Satellite Connectivity

Data transmits via low-Earth-orbit satellites — Starlink or equivalent. Latency is higher than a terrestrial fibre connection but entirely adequate for inference workloads. The ocean platform talks to the world without touching it.

Why Now

The terrestrial data centre
is hitting a wall.

The constraint on AI growth in 2026 is not intelligence — it is power. More than half of the data centres planned to open in 2026 have been severely delayed or cancelled. The reasons are consistent: limited grid capacity, cooling water scarcity, permitting delays, community opposition, and the simple fact that building a gigawatt-scale facility takes years and billions of dollars before a single computation runs.

AI demand is not waiting for the grid to catch up. The hyperscalers — Microsoft, Google, Amazon, Meta — are spending $300 billion annually on data centre infrastructure and still falling behind demand. The constraint is real, it is urgent, and it is structural.

"There are three sources of energy on the planet with tens of terawatts of new capacity potential: solar, nuclear, and the open ocean. We've built a technology platform that operates in the planet's most energy-dense wave regions, far from shore." — Garth Sheldon-Coulson, CEO, Panthalassa

Panthalassa's CEO is not wrong. The northern Pacific Ocean has consistent, powerful wave energy that dwarfs any terrestrial renewable resource in energy density per square kilometre. And unlike solar, it runs 24 hours a day regardless of weather. Unlike nuclear, it can be deployed in months not decades. Unlike grid-connected terrestrial data centres, it requires no community approval and no permitting process.

The Thiel Read

What this investment
actually says.

Reading the Thiel Signal

Thiel's investment philosophy has one consistent characteristic: he backs things that look like science fiction but are technically already real. The pattern from Facebook to Palantir to SpaceX is identical — the technology exists, the market doesn't believe it yet, and the window to invest before belief arrives is closing.

Panthalassa has been running prototypes since 2021. Ocean-1, Ocean-2, and the Wavehopper have all been deployed in sea trials. The $140 million Series B is not funding the science — the science is done. It is funding the manufacturing facility in Portland and the deployment of Ocean-3 in the Pacific later this year. This is a commercialisation round, not a research round.

That distinction matters enormously. When Thiel leads a commercialisation round, it means he has evaluated the prototype data and concluded the technology is ready to scale. That is a different signal from backing a research-stage company. The ocean compute is not a concept. It is a product entering production.

The NGE Connection
NGE Long-Horizon View

What this means
for the investor.

NGE Verdict · Panthalassa & Ocean Compute
🟢
The problem is real and urgent. Data centre capacity is the binding constraint on AI growth. Grid limitations are structural not cyclical. The demand for alternative compute infrastructure is not speculative — it is a current crisis.
🟢
The technology has been validated. Five years of prototype deployment. Thiel, Doerr, and Benioff in the same round. This is not a paper thesis — it is a product entering commercial deployment in 2027.
🟢
The unit economics are compelling. Zero fuel cost. Near-zero cooling cost. No land cost. No permitting cost. If the platform works at scale, the cost per computation could be dramatically lower than terrestrial alternatives.
🟡
The execution risk is real. Operating autonomous hardware in the open Pacific Ocean is genuinely hard. Saltwater corrosion, storm survivability, satellite latency, and manufacturing at scale are all unsolved at commercial volumes. The technology works in trials. Commercial scale is a different test.
🟡
Access is private for now. Panthalassa is pre-IPO. Direct investment requires accreditation and secondary market access. The public play is indirect — wave energy ETFs, satellite connectivity companies, and the hyperscalers who will be Panthalassa's first customers.
🔵
Watch 2027 closely. The commercial deployment is the proof point. If Ocean-3 delivers reliable AI inference from the Pacific at competitive cost, the valuation re-rates dramatically and the IPO timeline accelerates. 2027 is when the thesis becomes a fact.

The ocean is the largest untapped energy resource on earth. The AI economy is the largest unsatisfied energy consumer in history. Panthalassa is building the bridge between them. Peter Thiel has seen the prototype. He put $140 million on it. That is not nothing. That is a signal.

NGE · A Futuristic Investment Letter · Issue 05

Written from first principles. Not consensus. Not noise. Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Published when something is worth saying — not on a schedule.

— Pawan Bhatia · NextGen Economics · Bangalore, India · June 2026