Southern California Edison is reducing grid interconnection application evaluation times from 30-45 days to just two minutes, thanks to software from ThinkLabs AI. The company is using the NVIDIA CUDA platform to build digital twins and agents that proactively address grid instability by modeling scenarios shifting from static grid views to risk-informed investment strategies. NVIDIA is highlighting ThinkLabs, alongside four other companies, as innovators in using AI to accelerate the transition to a low-carbon energy future.
ThinkLabs AI Accelerates Grid Interconnection with Digital Twins
ThinkLabs AI has reduced grid interconnection application evaluation times to two minutes, a dramatic reduction from the typical 30-45 day process, by deploying AI-powered simulations. This approach allows for optimized investment strategies and a more resilient energy infrastructure as renewable sources introduce greater variability.
Josh Wong, founder and CEO of ThinkLabs, emphasizes the dual benefit of their solution, stating, “That hits not just reliability objectives, but also affordability, so we know how to maximize and optimize investments.” The core of this acceleration lies in an AI agent that identifies solutions to interconnection barriers through rapid grid simulations, a process previously reliant on lengthy manual reviews. This agent doesn’t simply assess current grid capacity; it models potential future scenarios, allowing utilities to anticipate and mitigate challenges before they arise. The company’s focus extends beyond simply speeding up approvals, aiming to fundamentally reshape how grid investments are prioritized.
ThinkLabs’ methodology reflects a broader shift toward intelligent grid management, recognizing that traditional approaches are insufficient for integrating increasing amounts of intermittent renewable energy. “The grid is getting less and less certain, so the ability to see things not statically, but as a probability — hence all the utility actions — should be risk informed,” Wong explained.
This probabilistic modeling allows for a more nuanced understanding of grid vulnerabilities, enabling targeted upgrades and proactive maintenance. The company’s work aligns with a growing recognition that artificial intelligence is no longer a supplemental tool for the energy sector, but a foundational element for achieving a sustainable and reliable power supply.
Twenty years ago, our founders, including Bill Gates, realized that emissions avoidance should also be a part of this energy solution.
Chris Levesque, president and CEO of TerraPower
Atomic Canyon’s Neutron & NIVA Streamline Nuclear Plant Operations
Atomic Canyon is redefining nuclear plant operations through AI-driven knowledge management, using NVIDIA accelerated compute with its Neutron and NIVA platforms. Neutron functions as an AI workbench, converting decades of accumulated nuclear procedures, regulatory guidance, and operational data into a searchable knowledge layer for nuclear professionals. Complementing this, NIVA, the Nuclear Industry Virtual Assistant, provides AI-powered capabilities across the national nuclear fleet, developed in collaboration with the Institute of Nuclear Power Operations, Electric Power Research Institute, and the Nuclear Energy Institute.
The impetus for this technological overhaul stems from the need to modernize a historically stable, yet increasingly challenged, sector. “Nuclear is a known technology we’ve been doing for 50 to 60 years, but the way we’ve been doing things simply will not scale to meet the moment that’s in front of us; it needs to be reinvigorated by artificial intelligence,” explained Trey Lauderdale, founder and CEO of Atomic Canyon.
This revitalization isn’t simply about automating existing processes, but fundamentally altering how nuclear facilities are managed and maintained. The platforms aim to distill complex information, making it readily accessible to operators and reducing the risk of human error in critical situations. TerraPower, another company highlighted for its advancements, is focused on emissions avoidance with its Natrium reactor, building on a vision established two decades ago.
Chris Levesque, president and CEO of TerraPower, stated, “Twenty years ago, our founders, including Bill Gates, realized that emissions avoidance should also be a part of this energy solution.” Meanwhile, Commonwealth Fusion Systems (CFS) is prioritizing inherent safety in its fusion reactor design.
Brandon Sorbom, cofounder and chief science officer of CFS, emphasized the passive safety features, noting, “It is passively safe, since the default mechanism is shutting itself down.” CFS utilizes high-temperature superconductors to create stronger magnetic fields, enabling a significantly smaller and more cost-effective reactor design, with the first ARC power plant planned for Virginia and grid connection anticipated in the 2030s. Sorbom projects that this plant “will be able to build a first-of-kind plant that will be cost-competitive with both renewable and nonrenewable sources of energy.”
Using these batteries with our power electronics and systems control software, you can make a highly responsive power source that can deal with the novel fluctuations of AI training.
Colin Campbell, chief technology officer of Redwood Materials, another NVIDIA Inception member
Redwood Materials Repurposes EV Batteries with NVIDIA Blackwell AI
Redwood Materials is deploying large-scale energy storage solutions for artificial intelligence facilities by repurposing 100%-recycled electric vehicle batteries, significantly reducing the time to bring new capacity online. This integration allows for a secure and flexible energy supply for the rapidly increasing electricity demands of AI training and deployment.
The system’s architecture simplifies power delivery by minimizing the need for transformers and uninterruptible power supplies, and enabling the use of either repurposed or new EV batteries, lowering overall costs, NVIDIA says. Redwood’s approach addresses a critical bottleneck in the AI industry, where electricity demand is currently outpacing the development of new grid infrastructure.
Colin Campbell, chief technology officer of Redwood Materials, explained, “Using these batteries with our power electronics and systems control software, you can make a highly responsive power source that can deal with the novel fluctuations of AI training.” This responsiveness is key to maintaining stable power for computationally intensive AI workloads. Beyond serving AI factories, Redwood’s technology is designed to contribute to the broader national grid, offering a scalable solution for energy storage and grid stabilization. The AI-powered intelligence layer continuously monitors and adapts to real-time energy needs, optimizing performance and ensuring a reliable power supply.
By combining battery engineering, power electronics, and advanced AI, Redwood Materials is positioning itself as a key enabler of sustainable growth in the AI era.
Nuclear is a known technology we’ve been doing for 50 to 60 years, but the way we’ve been doing things simply will not scale to meet the moment that’s in front of us; i t needs to be reinvigorated by artificial intelligence.
Trey Lauderdale, founder and CEO of Atomic Canyon
Commonwealth Fusion Systems Uses Omniverse for Fusion Energy Replication
Commonwealth Fusion Systems is dramatically accelerating the development of its SPARC tokamak through the use of NVIDIA Omniverse libraries and OpenUSD, compressing years of experimentation into weeks. This digital replication allows the team to model and refine the complex physics of fusion before physical construction, a process for achieving viable fusion energy, according to NVIDIA. The company uses this technology to simulate scenarios within the SPARC machine, a demonstration device designed to replicate the sun’s power source on Earth, and optimize its design for maximum efficiency and stability.
High-temperature superconductors are central to CFS’s approach, enabling magnetic fields significantly stronger than those used in previous fusion systems. These stronger fields allow SPARC to be 40 times smaller than conventional designs, reducing both construction and operational costs. The impact of this size reduction extends to CFS’s planned ARC power plant, slated for construction in Chesterfield County, Virginia.
Brandon Sorbom projects that ARC will be more than 10 times smaller than existing nuclear facilities, further driving down costs and accelerating deployment. The integration of NVIDIA’s Omniverse platform is a core component of CFS’s accelerated development timeline, enabling rapid iteration and optimization of the ARC plant’s design before groundbreaking.
The grid is getting less and less certain, so the ability to see things not statically, but as a probability – hence all the utility actions – should be risk informed.
Josh Wong, founder and CEO of ThinkLabs, a member of the NVIDIA Inception program for cutting-edge st
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
