NVIDIA Highlights Five Clean-Energy AI Projects at Climate Week

A Southern California Edison grid-review result is the clearest operational outcome in a broader showcase spanning nuclear operations, battery storage and long-range reactor development.

By 2 min read
NVIDIA Highlights Five Clean-Energy AI Projects at Climate Week
NVIDIA Highlights Five Clean-Energy AI Projects at Climate Week

Listen to this story

The audio brief

About 1:40
0:001:40
Read transcript
Southern California Edison says it cut the time needed to evaluate a grid-interconnection application from 30 to 45 days down to just two minutes, using software from ThinkLabs AI. That is the clearest operational result in NVIDIA’s clean-energy showcase at New York Climate Week. NVIDIA says a ThinkLabs agent ran grid simulations and identified ways around interconnection barriers. The company does not provide independent validation or a detailed methodology, so the result is notable but not yet easy to compare or reproduce. The broader showcase spans five companies: ThinkLabs AI, Atomic Canyon, Redwood Materials, TerraPower, and Commonwealth Fusion Systems. Their work covers everything from grid access to nuclear operations and fusion development. Redwood is proposing off-grid power systems for AI factories, combining recycled electric-vehicle batteries, power electronics, controls, and NVIDIA’s Blackwell computing platform. Redwood says an AI control layer could adapt power delivery to a data center’s changing demand, while using fewer transformers and inverters and no uninterruptible power supply. But those claims describe a system in development, not a reported deployment result. The longest-range project comes from Commonwealth Fusion Systems. It targets grid connection for its first ARC fusion plant in Chesterfield County, Virginia, during the 2030s, and says its SPARC design uses superconducting magnets to be 40 times smaller than earlier fusion systems. The key question is whether these demonstrations mature into independently measured, deployed infrastructure.

Story brief

3 key points

NVIDIA is using New York Climate Week to showcase energy applications spanning grid interconnection, battery-backed power systems, nuclear, and fusion. The strongest operational datapoint comes from Southern California Edison: NVIDIA says ThinkLabs AI reduced application-evaluation time from 30–45 days to two minutes, though the article provides no independent validation or methodology. Redwood Materials is...

  1. 01

    Southern California Edison’s reported evaluation time fell from 30–45 days to two minutes.

  2. 02

    NVIDIA says a ThinkLabs agent used grid simulations to identify interconnection solutions.

  3. 03

    Redwood’s proposed system combines recycled EV batteries, power electronics, controls, and NVIDIA Blackwell.

NVIDIA is highlighting five clean-energy companies at New York Climate Week. Its most concrete example comes from Southern California Edison, which used ThinkLabs software to cut evaluation time for each grid-interconnection application from 30–45 days to two minutes.

The companies work on different parts of the power system. NVIDIA named ThinkLabs AI, Atomic Canyon, Redwood Materials, TerraPower and Commonwealth Fusion Systems as users of its technology. Their projects range from connecting resources to the grid to designing nuclear and fusion plants.

The grid-review result

From reactor records to onsite power

Redwood Materials is developing off-grid power systems for AI factories with recycled electric-vehicle batteries, power electronics and control software. Redwood says an AI layer running on NVIDIA Blackwell can adjust the system to a data center’s changing energy needs. The company also says the design requires fewer transformers and inverters, with no uninterruptible power supply.

Equipment associated with Redwood Materials’ battery-based power approach for AI factories.
Redwood Materials is developing off-grid AI-factory power systems using recycled EV batteries and an AI control layer. Source: blogs.nvidia.com.

Development tools with longer horizons

CFS plans its first ARC fusion plant in Chesterfield County, Virginia, with grid connection targeted for the 2030s. It says superconducting magnets make SPARC 40 times smaller than earlier fusion systems. Unlike Edison’s reported operational result, these examples describe tools, objectives and future plans rather than directly comparable published outcomes.

Sources

  1. blogs.nvidia.com5 Companies Using NVIDIA AI for Clean Energy

Loading discussion...

YOUR READING SPACE

Notifications