As data-center developers frantically seek to secure power for their operations, one startup is proposing a novel solution: Build them into floating offshore wind turbines.

San Francisco–based offshore wind-power developer Aikido Technologies today announced its plans to start housing data centers in the underwater tanks that keep its turbine platforms afloat. The turbines will supply the power for the servers, and onboard batteries and grid connection will provide backup.

The company’s first prototype, a 100-kilowatt unit, is scheduled to launch in the North Sea off the coast of Norway by the end of this year. A 15-to-18-megawatt project off the coast of the United Kingdom may follow in 2028.

Aikido is one of several companies planning data centers in unusual places—underwater, on floating buoys, in coal mines and now on offshore wind turbines. The creativity stems from the forces of several trends: rapidly rising energy demand from data centers, the need for domestic renewable power production, and limited real estate.

The North Sea serves as an ideal first spot for floating, wind-powered data centers because European policymakers and companies are looking to regain domestic control over energy production. They’re also looking to host an AI economy on servers within the continent’s boundaries. Floating wind platforms keep the compute out of sight while tapping the stronger, more consistent air streams that blow over deep waters, where traditional, seabed-mounted turbine monopiles can’t go.

“A lot of energy in the clean-energy space is focused on powering AI data centers quickly, reliably, and cleanly in a way that does not upset neighbors and remains safe, fast, and cheap,” says Ramez Naam, an independent clean-energy investor who does not have a stake in Aikido. “Aikido has that, and a smart team,” he says.

Floating Wind-Power Designs Evolve

Aikido’s design builds on many iterations tested by the growing floating wind industry. When Norwegian energy giant Equinor finished construction on the world’s first floating wind farm in 2017, it kept the turbines upright with ballasted steel columns extending 78 meters into the water—a design called a spar platform. This gave it a dense mass like the keel of a boat. Since then, the floating wind industry has largely coalesced around a semisubmersible design based on oil and gas platforms. Semisubmersibles don’t go as deep as spar platforms; instead, they extend buoyancy horizontally. Anchors, chains, and ropes keep the platform floating within a certain radius.

Aikido is taking the semisubmersible approach. Its football-field-size platform holds the turbine in the center, and three legs extend tripod-like outward, like a Christmas-tree stand. At the end of each leg is a ballast that reaches 20 meters deep. This holds tanks largely filled with fresh water to maintain the platform’s buoyancy in the salty ocean.

The data centers will go in the upper part of each ballast tank. There’s room for a 3- to 4-MW data hall in each tank, giving the platform a combined compute of 10 to 12 MW. Below the data halls is an open chamber used as a safety barrier, and below that sit the freshwater tanks. The water is piped up to the data center for liquid cooling of the servers. The warmed water is then funneled back down the ballast into the tank. There, proximity to the cold ocean water cools it again as the heat is conducted out through the tank’s steel walls.

“We have this power from the wind. We have free cooling. We think we can be quite cost competitive compared to conventional data-center solutions,” says Aikido CEO Sam Kanner. “This crunch in the next five years is an opportunity for us to prove this out and supply AI compute where it’s needed.”

One challenge, he says, is that liquid cooling can’t cover all the data center’s needs. For example, heat generated from Ethernet switches that connect the GPUs can’t be liquid-cooled with commercially available technology. So Aikido installed an air-conditioning method for that.

Another challenge is the marine environment, which is “pretty brutal to engineer around because there’s the increased salinity, there’s debris, and there’s various kinds of corrosion and fouling of metal piping that you wouldn’t have in a freshwater environment,” says Daniel King, a research fellow at the Foundation for American Innovation in Washington who focuses on AI infrastructure.

Offshore Data Centers Face Challenges

Aikido’s plan avoids the prickly not-in-my-backyard complaints that are dogging both onshore wind and data-center projects. It might also circumvent some inquiries into water usage and power demand too, or so Aikido’s thinking goes.

But it might not be that easy. “Instinctively many people reach for offshore or even orbital outer-space data centers as a way to circumvent the typical burdens of environmental reviews,” says King. “But there could be more or additional requirements around discharging heat and the effects that has on marine life that are different from the considerations of a terrestrial data center. It’s unclear to me whether this actually makes life easier or harder for a developer.”

Prefabricated data halls could be installed quayside, followed by final electrical and plumbing connections to commission the data center.Aikido

Aikido’s “design choice to use the fresh water in the ballast as a working fluid is a novel one” that, thanks to the closed-loop system, may “alleviate some of the engineering problems you see when a really high temperature fluid is pumping its heat directly into a marine environment,” King says.

Offshore sites are also vulnerable to sabotage, King notes. Since Russia’s invasion of Ukraine, fleets of vessels directed by the Kremlin have reportedly started messing with offshore wind and communications infrastructure in northern Europe. Russian and Chinese boats have allegedly cut subsea cables in recent years.

But vandalism is a risk anywhere, including at conventional data centers, Aikido CEO Kanner notes. Unlike those on land, where the local police have jurisdiction, Aikido’s data centers would enjoy protection from national coast guards, which he suggests gives an added degree of security.

North Sea Hosts Clean Energy

Kanner first began thinking about offshore wind turbines as a place to build data centers after a chance phone call with a cryptocurrency billionaire. The financier wanted to know whether turbines in international waters could power servers generating digital tokens at a moment when crypto-mining faced increased scrutiny from regulators. The talks fizzled. But that encounter sparked Kanner’s curiosity about how to use power generated onboard floating turbines.

When ChatGPT emerged in 2022 and sparked a heated debate over how to power and cool such technology, the idea to put the data center in the floating turbine clicked for Kanner. The idea really congealed after he met with the chief executive of Portland, Ore.–based Panthalassa. The wave-energy company was proposing to enclose small, remote data centers in buoys attached to equipment that generates power from the surf. Panthalassa just completed its full-scale prototype tests off the coast of Washington state last summer.

At that point, Aikido had already designed a modular platform for floating wind turbines. Each platform consists of 13 major steel components that are snapped together with pin joints—like IKEA furniture. The platforms fold up in a flat configuration that takes up roughly half the space of other designs, allowing it to be transported by a wider range of ships, according to Aikido. From there, it was a matter of figuring out how to accommodate a data center in the unused space.

Aikido’s prototype will use a refurbished Vesta V-17 turbine. It will need onboard batteries for backup power and will also be connected to the grid for additional power during seasons with less wind. Aikido envisions eventually sprinkling its data centers among large arrays of offshore turbines to tap into that larger power infrastructure.

Between Russia’s threat to expand its war in Ukraine to EU countries and the Trump administration’s bid to pressure Denmark into ceding sovereignty of Greenland to Washington, Europe is scrambling to build up its own energy production and AI capabilities. The North Sea, increasingly, looks like a primary theater of that effort. In January, nearly a dozen European nations banded together in a pact to transform the North Sea into a “reservoir” of clean power from offshore wind.

In 1950, the English mathematician Alan Turing devised what he called “the imitation game.” Later dubbed the Turing test, the experiment asks a human participant to conduct a conversation with an unknown partner and try to determine if it’s a computer or a person on the other end of the line. If the person can’t figure it out, the machine passes the Turing test.

Power grid operators are now preparing for their own version of the game. Virtual power plants, which concatenate small, distributed energy resources, are increasingly being tapped to balance electricity supply and demand. The question is: Can they do their job as well as conventional power plants?

Grid operators can now find out by running these power plants through a Turing-like test called the Huels. To pass the Huels test, the performance of a virtual power plant must be indistinguishable from that of a conventional power plant. A human grid operator serves as the judge.

Virtual power plant developer EnergyHub, based in Brooklyn, N.Y., developed the test and outlined it in a white paper released today. “What we’re really trying to do is fool the operators into feeling that these virtual power plants can act and feel and smell like conventional power plants,” says Paul Hines, chief scientist at EnergyHub. “This is a kind of first litmus test.”

What Are Virtual Power Plants (VPPs)?

The virtual-versus-conventional power plant question is a timely one. Virtual power plants, or VPPs, are networks of devices such as rooftop solar panels, home batteries, and smart thermostats that come together through software to collectively supply or conserve electricity.

Unlike conventional power generation systems, which might crank up one big gas plant when electricity demand peaks, VPPs tap into small, widely disbursed equipment. For example, a VPP might harness electricity from hundreds of plugged-in electric vehicles or rooftop solar panels. Or it might direct smart thermostats in homes or businesses to turn down heat or cooling systems to reduce demand.

The technology is emerging at a time when concerns over data centers’ electricity demand is hitting a fever pitch. The consultancy BloombergNEF estimates data-center energy demand in the United States will reach 106 gigawatts by 2035–a 36 percent jump from what it had projected just seven months ago.

How utilities and grid operators will meet the growing demand is unclear and faces challenges on many fronts. Turbines for natural gas plants are backordered, and new nuclear reactors are still years away. Wind and solar, while cheap and fast to build, don’t produce the 24/7 electricity that data centers demand and face an uphill political battle under the Trump administration.

All of this together has created an opening for VPPs, which could add gigawatts to the grid without significantly jacking up electricity rates. “It’s a political issue. If you said you’re going to get electricity costs under control, this is literally the only way to do it in 12 months,” says Jigar Shah, a clean energy investor at Multiplier in Washington, D.C., who led the U.S. Department of Energy’s Loan Programs Office under the Biden administration.

VPPs could also reduce utilities’ need to invest in distribution equipment, avoiding supply chain shortages and inflated costs, Shah says. “There is no other idea that you could possibly deploy in 12 months that would have that big of an impact,” he says.

According to a 2024 U.S. Department of Energy report, VPPs could provide between 80 and 160 gigawatts of capacity across the U.S. by 2030—enough to meet between 10 and 20 percent of peak grid demand.

How Can VPPs Gain Grid Operator Trust?

But first, VPP developers have to win over grid developers. Benchmarks like the Huels test are crucial to building that trust. “In order for us to build our reliance on VPPs, they do need to pass the Huels test, and operators need to be able to count on” the VPPs delivering power when called upon, said Lauren Shwisberg, a principal in the nonprofit research group Rocky Mountain Institute who co-authored a recent report on VPPs and was not involved in the development of the test.

Matthias Huels, an engineer who spent more than four years at EnergyHub, first came up with the idea for the test in 2024. After workshopping the idea with colleagues and, somewhat ironically, ChatGPT, Huels presented the concept to the company.

Huels designed the test subjectively. Currently, in its earliest iteration, it appears to follow a guideline akin to the Supreme Court’s “I know it when I see it” test for what distinguishes pornography from erotic art. That is to say: Passing the test depends on who’s judging. If a grid operator finds the power from a VPP as dependable as electricity from an actual power plant burning gas to produce electrons, then the VPP has passed.

There are four levels to the Huels test. To reach level 1, a VPP must be able to shave off demand from the grid by, for example, successfully scheduling smart thermostats to dial down when the grid faces maximum demand. To reach level 2, a VPP must be able to respond to market and grid data and dial down demand when prices hit a certain level or tap into solar panels or batteries when power is needed. Human decision makers are involved at these levels.

Passing the Huels test comes at level 3. That’s when a VPP can function automatically because it’s proven reliable enough to be indistinguishable from a gas peaker plant–the type of power station that comes online as backup only when the grid is under stress. Passing level 4 involves VPPs acting fully autonomously to adjust output based on a number of actively changing variables throughout the day.

“The imitation game that Alan Turing came up with was: Can a computer fool an interrogator to think it’s actually human even though it’s a computer,” Hines says. “We propose this idea of a test that would allow us to say: Can we fool a grid operator into thinking that the thing that’s actually solving their problems is this aggregation of many devices instead of a big gas plant?”

Can VPPs Mimic Gas Peaker Plants?

Peaker plants only generate power about 5 percent of the time over their lifespans. That makes them easier for VPPs to mimic because, like peaker plants, the limited amount of power that can be made available by demand response or harvested from batteries only provides bursts of power that last a few hours at a time.

Far more difficult is stacking up to a full-scale gas plant, which operates 65 percent of the time or more, or a nuclear plant, which usually operates at least 95 percent of the time. Getting there would involve equipping a VPP network with long-duration storage that could be powered up during the day when solar panels are at peak output and discharged all night long. “You start talking about VPPs with large amounts of batteries that can run 365 days per year,” Hines says. “That’s a road we can go down.”

EnergyHub has been putting its VPP systems through the Huels test. Last year, EnergyHub successfully ran trials with Arizona Public Service, Duke Energy in North Carolina, and National Grid in Massachusetts. In Arizona, EnergyHub’s software dialed into homes with solar panels and smart thermostats and ran air conditioners to “pre-cool” houses during the day when the sun was generating lots of electricity. This allowed the state’s biggest utility to reduce demand during peak hours when residents would typically return home from work to turn on televisions and crank up their air conditioners.

“You have too much power in the middle of the day because of solar, then the early evening comes and you get people ramping up their evening loads right as the solar is ramping down,” Hines says. “You need something that can feather through that schedule. We created something that can do this.”

That lands the company somewhere between a 2 and 3 on the Huels testing scale. Passing level 3 “is going to take a few years,” Hines says.