How AI Is Accelerating Clean Energy: 5 Companies Building the Grid of the Future

Clean energy is not difficult to obtain, but outdated infrastructure, lengthy research and development timelines, and high upfront costs have historically slowed large-scale adoption.

At New York Climate Week, NVIDIA is highlighting five companies using artificial intelligence to advance clean energy projects. Their technologies are helping accelerate the path from research to commercialization while supporting the development of a low-carbon power grid.

1. ThinkLabs AI uses digital twins to accelerate the autonomous grid

ThinkLabs AI develops digital twins and AI agents designed to help utilities manage an increasingly complex power system. Built on the NVIDIA CUDA Platform, the technology accelerates grid interconnection timelines, integrates clean energy sources, and optimizes the power grid to manage ongoing fluctuations.

“As the grid becomes less and less certain, the ability to see things as probabilities rather than statics, and therefore every action taken by utilities, needs to be risk-informed,” said Josh Wong, founder and CEO of ThinkLab and a member of the NVIDIA Inception program for cutting-edge startups.

“We know how to maximize and optimize our investment because this not only meets reliability goals, but also affordability goals.”

Southern California Edison used ThinkLabs software to reduce the time required to evaluate each grid interconnection application from 30 to 45 days to just two minutes.

The improvement comes from ThinkLabs agents that run grid simulations and identify solutions to interconnection barriers.

2. Atomic Canyon applies AI to nuclear power plant operations

Atomic Canyon, another NVIDIA Inception member, uses the Neutron and NIVA platforms, powered by NVIDIA Accelerated Computing, to provide AI-powered knowledge management and reactor operations assistance for nuclear power plants.

Neutron is an AI workbench built for nuclear professionals. It transforms procedures, regulatory guidance, operational experience, license records, design calculations, and structured and unstructured data into a unified knowledge layer.

NIVA, which stands for Nuclear Industry Virtual Assistant, was developed in collaboration with the Nuclear Operations Research Institute, the Electric Power Research Institute, and the Nuclear Energy Research Institute. The integrated platform brings AI-powered capabilities to national nuclear fleets.

“Nuclear power is a known technology that we’ve been working on for 50 to 60 years, but the way we’ve always done it doesn’t scale up to the moment in front of us. It needs to be reinvigorated by artificial intelligence,” said Trey Lauderdale, founder and CEO of Atomic Canyon.

3. Redwood Materials repurposes EV batteries to power AI data centers

AI-driven electricity demand is increasing faster than new grid infrastructure can be built, creating a bottleneck for industry growth.

Redwood Materials is addressing that gap by combining its expertise in battery engineering, power electronics, and system design to build large-scale, off-grid power solutions for AI factories. The approach can bring new capacity online in months rather than years.

Redwood’s 100% recycled electric vehicle batteries provide on-site energy storage for electricity providers. They create a secure and flexible energy system for AI factories and the broader national grid.

The batteries are integrated into existing clean energy infrastructure with an AI intelligence layer. The NVIDIA Blackwell Platform enables these systems to remain highly responsive and adapt to the real-time energy demands of the data centers they power.

This approach also reduces power costs. Redwood’s simplified architecture uses in-house power electronics, requires far fewer transformers and inverters, eliminates the need for uninterruptible power supplies, and supports the reuse of existing or new electric vehicle batteries.

“When used in conjunction with our power electronics and system control software, these batteries create a responsive power source that can handle the emerging variability of AI training,” said Colin Campbell, chief technology officer at Redwood Materials and another NVIDIA Inception member.

4. TerraPower develops sodium reactors for safer, sustainable nuclear power

Nuclear power currently provides 20% of the country’s electricity. TerraPower is working to make more electricity available safely, sustainably, and quickly through emissions-free sodium reactors.

“Twenty years ago, our founders, including Bill Gates, recognized that emissions reduction had to be part of this energy solution,” said Chris Levesque, president and CEO of TerraPower.

TerraPower is using NVIDIA Omniverse, a platform for creating digital twin software, to support efforts to accelerate future plant siting and delivery from years to months.

Sodium reactors use a design that differs from traditional nuclear power generation. Because liquid metallic sodium cools the reactor instead of water, the system can operate at lower pressures comparable to atmospheric pressure.

The reactor does not require off-site water or electricity to maintain stability. In an emergency, it can be cooled without human intervention.

5. Commonwealth Fusion Systems works toward grid-ready fusion energy

Commonwealth Fusion Systems (CFS) is working to bring fusion energy into the clean energy mix in the 2030s.

Using the NVIDIA Omniverse library and OpenUSD, CFS is compressing years of experimentation into weeks with SPARC, its demonstration fusion machine. SPARC successfully replicated the sun’s power source on Earth.

Fusion energy is a safe, carbon-free source of electricity.

“With fusion, you can’t get out of control,” said Brandon Thorbom, CFS co-founder and chief scientific officer. “It’s passively safe because the default mechanism shuts it down automatically.”

By using high-temperature superconductors, CFS can create stronger magnetic fields than previous fusion systems.

Those magnets make SPARC designs 40 times smaller, which makes them cheaper to build and operate. CFS’s ARC plants, the first of which will be built in Chesterfield County, Virginia, and connected to the grid in the 2030s, are more than 10 times smaller.

“We will be able to build a first-of-its-kind plant that is cost-competitive with both renewable and non-renewable energy sources,” Sobom said.

See more details in Sustainability use cases powered by NVIDIA.