Sam Atlman-backed company Exowatt wants to power AI data centers with billions of hot rocks

When Hanan Habi started thinking about how to solve the AI ​​energy crisis, he kept one number in mind: one cent per kilowatt-hour.

“We went through all kinds of configurations and designs,” Happy, Exowatt co-founder and CEO, told TechCrunch. “They all look different from each other. We tried to learn from everyone: How can I reduce structural costs? How can I reduce maintenance costs? How can I improve this?”

After years of brainstorming and building, Exowatt’s first step toward achieving this goal was a simple shipping container-sized box topped with a transparent canopy. The interior is similarly simple. If Exowatt can deliver on its promise of cheap solar power that generates electricity 24/7, it could turn the data center market and the broader energy world upside down, providing power around the clock at a very low cost.

To scale production in pursuit of the 1 cent per kilowatt-hour target, Exowatt has raised an additional $50 million in an extension of a $70 million Series A round that closed in April, TechCrunch has learned exclusively.

The expansion was led by MVP Ventures and 8090 Industries with participation from Atomic, BAM, Bay Bridge Ventures, DeepWork Capital, Dragon Global, Florida Opportunity Fund, Massive VC, New Atlas Capital, Overmatch, Protagonist, and StepStone. Previous investors include Andreessen Horowitz and Sam Altman.

Happe said Exowatt wasn’t looking to raise additional capital after the April round, but the “strong momentum we’ve seen in the market” and “strong investor interest” encouraged him to take on the new funds at a higher valuation.

Exowatt’s backlog currently stands at about 10 million P3 units representing 90 gigawatt-hours of capacity, he said. “The goal is to expand as quickly as possible to millions of units and eventually billions of units,” he said. Happy said the company should meet its 1-cent target when production reaches about 1 million units per year.

Exowatt is essentially repackaging technology that has been around for decades. It is known as concentrated solar energy or solar thermal energy, and uses the sun’s energy to heat materials that are good for storing or transmitting thermal energy. In cases where this thermal energy is stored for a long period of time, these materials tend to be derived from or resemble rocks – hence the technology’s nickname “rocks in a box.”

Each P3 consists of a metal box topped with lenses that focus sunlight into a narrow beam. This beam then heats a special brick inside the shipping container. A fan blows air over the bricks to transfer the heat to another box containing a Stirling engine (a piston-powered device that converts heat into mechanical energy) and a generator. To store more energy, developers will install more P3 boxes. “Everything is designed to be very simple,” Happy said.

Each thermal battery can retain heat for up to five days, ensuring continuous operation, and Exowatt will connect several of them together to feed a single generator unit. The number depends on how quickly and how much electricity the customer wants to generate. Happi said the system’s efficiency is on par with solar photovoltaic panels, and slightly better than photovoltaics coupled with lithium-ion batteries.

Other companies have built different approaches to the same technology, though most have failed to compete with solar PV and lithium-ion batteries, both of which surprised experts by how quickly the cost came down.

Happe argued that P3’s small size and Exowatt’s iterative approach sets it apart. He said there are just over 100 solar thermal or concentrated solar power projects planned, built or decommissioned worldwide. “If you compare that to the fact that we produce 1.5 billion solar panels a year, you can see that the learning curve effects are very far apart.

“What Exowatt is about is taking a modular system that we know works in principle, scaling up its manufacturing and then applying the manufacturing learning curves.”

Exowatt is unlikely to be cost effective everywhere, and the number of P3 modules needed to operate a data center would require huge amounts of land. Additionally, it does best in sunny areas, which may limit its broader impact.

But Happe sees “huge overlap” between where Exowatt P3 excels and where new data centers are being built. “We are not short of any projects to do,” he added.