Artificial intelligence is rapidly expanding, along with the energy requirements needed to sustain its operations. Modern AI data centers demand significantly more electricity than traditional cloud servers, leading to increased strain on existing power grids. One promising solution is the repurposing of electric vehicle (EV) batteries for AI data centers. Redwood Materials, established by Tesla co-founder JB Straubel, is at the forefront of this initiative, repurposing retired EV batteries into modular storage systems that can supply energy for AI operations. This approach not only minimizes environmental impact by reducing landfill waste and the need for new battery production but also offers cost-effective energy storage solutions, making it a viable and sustainable option for the growing AI industry.
Redwood Energy, a new venture by Redwood Materials, is addressing the energy needs of AI through innovative solutions. Instead of sending used electric vehicle batteries to recycling, the company gives them a second life by rebuilding them into modular storage systems that can power AI operations. The process begins with collecting and testing old EV battery packs, many of which retain over 50 percent of their original capacity. Once approved, these batteries are rebuilt into systems that can support AI data centers, offering a sustainable and cost-effective energy storage solution.
One example of this initiative is a 12-megawatt, 63-megawatt-hour microgrid now supporting a 2,000-GPU data center operated by Crusoe in Nevada. This project represents the largest active deployment of second-life EV batteries, operating more affordably than systems built with new batteries. The environmental and financial benefits of this approach are significant. By using repurposed EV batteries, companies can avoid sending usable materials to landfills, reduce mining for new raw materials, and lower carbon emissions. Additionally, second-life battery systems typically cost less than brand-new lithium-ion options, making large-scale energy storage more accessible to AI developers and utilities.
Redwood Energy estimates that over 100,000 electric vehicles will be retired in the United States this year, providing a substantial potential for energy storage. The company already has more than 1 gigawatt-hour of second-life battery capacity in its development pipeline, with a goal to reach 5 gigawatt-hours next year. Larger energy projects, including 100-megawatt sites, are also planned. Each one is designed to support the increasing power demands of AI data centers using repurposed batteries instead of new supply chains.
Repurposed EV batteries for AI data centers are proving to be a smart, sustainable solution for the industry’s soaring energy demands. Redwood Energy is helping to bridge the gap between battery recovery and recycling, powering the future of AI while reducing waste and emissions. As more EVs retire and AI continues to grow, expect to see even more green-powered data centers on the horizon.
One question remains: is repurposing EV batteries the missing link in building greener, smarter data centers, or just another temporary fix masking a bigger problem? The answer may depend on how effectively this technology can be scaled and integrated into the broader energy and tech ecosystems. For now, Redwood Energy’s initiative offers a compelling case for sustainable and efficient energy use in the AI sector.