Rune Raises $40M to Power AI Data Centers with Wasted Solar Energy

Rune Secures $40 Million to Deploy Solar-Powered AI Data Centers

Rune, a startup incorporated as Liitto Technology Inc., closed a $40 million Series A funding round led by 1

1

1

1

2

2

. The company debuted its first product alongside the funding announcement: RELIC, or Renewable Energy Linked Intelligent Compute, a modular computing system designed to run AI workloads directly at solar farms using excess solar power that would otherwise be wasted.

Tapping Into Wasted Renewable Energy to Meet Growing Power Demands of AI

Solar plants waste up to 20% of the energy they generate due to grid limitations, according to 1

1

1

1

. During peak production hours, grids often cannot absorb all available solar power. The California grid alone wasted around 3.5 million megawatt-hours of solar in 2025—enough to power hundreds of thousands of homes. Rune's modular data centers capitalize on this unused capacity, providing a solution to the growing power demands of AI without requiring new power plants or grid infrastructure.

CEO William Layden explained that every solar plant is essentially a latent data center. "The power is already there, sitting idle while AI labs wait years for grid connections that may never come," 2

2

2

2

. The company has already deployed RELIC at a 200-megawatt solar facility in Texas and plans to expand to wind farms.

RELIC System Cuts AI Infrastructure Costs by 85%

Building traditional AI data centers requires massive investments in auxiliary infrastructure. OpenAI, for instance, expects to spend $4.2 billion on power transmission hardware for its Ohio data center buildout. Rune claims RELIC reduces infrastructure costs by 2

2

2

2

by eliminating the need for long-haul power transmission lines and transformers.

The system achieves these savings through a direct current approach. Solar panels generate direct current electricity, and computers run on direct current electricity. "Nothing connected them before Rune," 1

1

1

1

. Traditional data centers convert DC power from solar panels to AC for transmission, then back to DC for servers—a process requiring expensive transformers. RELIC's custom power electronics safely convert high-voltage solar electricity directly to the right voltage for servers, bypassing transformers entirely.

Modular Approach Eliminates Years-Long Supply Chain Delays and Permitting Delays

Rune positions itself as delivering the fastest time to compute in the industry. "We don't need to wait three years, four years, five years for new infrastructure to get built. We're using what we have today to bring on compute today," 1

1

1

1

. Traditional approaches face significant supply chain delays, with gas power plants experiencing four-year waits for transformers alone.

Each RELIC module measures 8-by-8 feet and ships with up to 1,024 graphics cards, cooling equipment, and custom power electronics. The modular data centers require no construction—Rune simply places modules on-site using a forklift. Setup takes approximately one hour, and customers can begin running AI workloads within 2

2

2

2

. The system also avoids lengthy grid interconnection processes, permitting delays, and community opposition that typically accompany new power plant construction.

What This Means for Cleaner Energy and AI Infrastructure

Rune's approach addresses multiple bottlenecks simultaneously. By connecting directly to solar farms, the company provides AI labs with immediate access to cleaner energy while helping utilities monetize power that would otherwise go unused. The simplified supply chain—free from transformer shortages and long-haul transmission requirements—makes the system more robust than traditional AI infrastructure.

The implications extend beyond cost and speed. As AI companies face mounting pressure to reduce their environmental footprint, Rune offers a path to scale compute capacity using renewable energy without adding strain to existing grids. The startup's plans to expand beyond solar to wind farms suggest the model could be replicated across different renewable energy sources, potentially transforming how the industry thinks about powering AI workloads in the coming years.