Articles tagged with "AI-data-centers"
Geothermal energy could beat nuclear, coal to satiate AI power hunger
A recent Stanford University study highlights enhanced geothermal systems (EGS) as a promising clean energy technology capable of addressing the growing global demand for reliable, low-cost power. Unlike traditional geothermal power, which is limited to volcanic regions, EGS can be implemented nearly anywhere by drilling boreholes 3 to 8 km deep, fracturing rock, and circulating fluid to generate continuous electricity year-round. The study suggests that integrating EGS to supply about 10% of a nation’s electricity could significantly reduce reliance on wind, solar, and battery storage—cutting wind capacity needs by 15%, solar by 12%, and battery storage by 28%—while using far less land and lowering energy costs by at least 60% compared to fossil fuels. EGS also offers advantages over nuclear power, providing a steady baseload without risks of meltdown, radioactive waste, or weapons proliferation, and can be deployed more quickly and cheaply. This makes it especially suitable for power-hungry, off-grid applications
energygeothermal-energyclean-energyenhanced-geothermal-systemsrenewable-energyenergy-storageAI-data-centersRedwood attracts Google for its $425M Series E as AI power needs rise
Redwood Materials, a battery recycling and cathode production startup founded by former Tesla CTO JB Straubel, has raised $425 million in its Series E funding round, attracting new investors including Google alongside existing backers like Nvidia’s NVentures, Capricorn, and Goldman Sachs. This round, led by venture firm Eclipse, has pushed Redwood’s total capital raised to $4.9 billion and reportedly values the company at over $6 billion. The fresh capital is fueling Redwood’s expansion into energy storage solutions aimed at powering AI data centers and large industrial sites, a strategic pivot reflecting the surging electricity demand driven by AI, manufacturing, and electrification. Originally focused on creating a circular supply chain by recycling battery scrap from electronics and extracting key materials like nickel and lithium, Redwood has broadened its operations to include cathode production and a new business line called Redwood Energy. This venture repurposes used EV batteries into micro-grids to supply power for data centers, leveraging the company’s extensive inventory—over
energybattery-recyclingenergy-storageAI-data-centerselectric-vehiclesmaterials-recoverygrid-scale-storageHyAxiom's David Alonso on fuel cells and time to power for AI data centers
HyAxiom, a Connecticut-based fuel cell manufacturer and part of the Doosan group, is addressing the growing power challenges faced by rapidly scaling AI data centers. As AI facilities expand from tens of megawatts to potentially gigawatt-scale operations, traditional electric grids and local utilities struggle to deliver sufficient capacity quickly. HyAxiom’s stationary fuel cells offer a solution by enabling on-site power generation that can be deployed rapidly—within about 12 months—bypassing grid delays. Their fuel cells also produce significantly lower emissions compared to conventional combustion-based power sources, addressing environmental concerns as data centers grow more power-dense. A key product in HyAxiom’s portfolio is the PureCell® Model 400, a modular, containerized fuel cell system that integrates fuel handling, electricity generation, and power conversion. Each unit delivers 460 kilowatts and can be scaled incrementally to meet specific data center demands, from small to very large capacities. This modularity and rapid deployment capability make fuel
energyfuel-cellsAI-data-centersgreen-hydrogenpower-infrastructuremodular-power-systemsemissions-reductionWorld-first practical cryogen-free superconducting motor unveiled
US startup Hinetics, part of the POETS center and funded by the US Department of Energy’s ARPA-E, unveiled the world’s first fully integrated, cryogen-free superconducting motor at CES 2026. This motor is a practical proof-of-concept demonstrator developed over three years, designed for aerospace and AI data center applications. Unlike traditional superconducting machines that require bulky external cryogenic systems, Hinetics’ motor features a self-contained design using an onboard cryocooler to maintain low temperatures without external liquid cryogen cooling. The current unit is a scaled-down version of a larger six-megawatt motor under construction, aiming to make superconducting technology commercially viable by reducing costs, particularly of superconducting tape, which has halved in price over the past three years. The motor offers significant advantages for AI data centers and aerospace, including rapid response to transient power demands due to its very low inductance and high efficiency of approximately 99.5 percent. This efficiency
energysuperconducting-motorcryogen-free-technologyelectric-motorpower-generationaerospace-applicationsAI-data-centersUSS Nimitz reactors may fuel AI data centers after retirement
The USS Nimitz, the U.S. Navy’s oldest active nuclear-powered aircraft carrier, has recently completed its final deployment and entered a lengthy decommissioning process. Central to this process is the challenge of dismantling its two nuclear reactors, which provided the ship with near-unlimited endurance at sea. The decommissioning is complex, costly—expected to exceed $1 billion—and will unfold over multiple phases spanning potentially decades. Preparations are already underway, with Huntington Ingalls Industries awarded a contract to begin inactivation and defueling of the reactors. In light of the Nimitz’s retirement, a novel proposal has emerged to repurpose retired naval nuclear reactors to power artificial intelligence (AI) data centers. Texas-based HGP Intelligent Energy LLC suggests using these reactors to supply 450 to 520 megawatts of electricity for AI workloads, potentially at Oak Ridge National Laboratory. This approach could be more cost-effective than building new civilian nuclear plants, leveraging existing naval reactor technology, supply chains,
energynuclear-reactorsAI-data-centersUSS-Nimitzenergy-recyclingnuclear-powerdecommissioningNickel-Zinc Battery Company Raises $30 Million - CleanTechnica
ZincFive, a company specializing in nickel-zinc (NiZn) battery solutions for immediate power applications, has successfully raised $30 million in an oversubscribed Series F funding round, bringing its total funding to $254 million since its founding in 2016. The new capital will support rapid scaling of ZincFive’s commercial and manufacturing operations to meet growing demand, particularly from AI data centers requiring reliable power backup. ZincFive claims to be a world leader in NiZn battery technology, having deployed or contracted 2 gigawatts (GW) of battery backup capacity, notable for this less common battery chemistry. The company highlights its nickel-zinc batteries as safer and more sustainable alternatives to lead-acid and lithium-ion options, offering superior power density, inherent safety (zero thermal-runaway risk), and a 96% recyclable design. ZincFive’s technology is engineered for the dynamic power needs of AI data centers, which represent a significant and rapidly expanding market due to the massive power demands of AI
energynickel-zinc-batteryAI-data-centersclean-technologybattery-technologyenergy-storagesustainable-energyWorld's first 800V immersion-cooled battery targets AI power demand
XING Mobility is unveiling the world’s first 800V immersion-cooled high-voltage DC backup battery system, called BBx800, designed specifically for AI data centers at CES 2026. This system leverages full immersion cooling by submerging each battery cell in an insulating fluid to maintain stable operating temperatures between 25°C and 27°C, reducing thermal runaway risk and enhancing stability under extreme loads. The compact module supports voltage configurations of ±400V or 800V, and a standard 20 OU rack can deliver peak outputs up to 1 MW for three minutes or 1.2 MW for 90 seconds, targeting short-duration backup during rapid power fluctuations and peak AI workloads. The development responds to the rising power demands of AI data centers, where per-rack power consumption has surged from around 100 kW to over 1 MW, challenging traditional 48V architectures due to current, thermal, and efficiency constraints. NVIDIA’s 2025 push for 800V
energybattery-technologyimmersion-coolingAI-data-centershigh-voltage-systemsenergy-storagethermal-managementState of Commercial Electrification — 2026 Outlook - CleanTechnica
The commercial electrification market is poised for significant transition in 2026 amid evolving regulatory frameworks and federal incentives under review. Original Equipment Manufacturers (OEMs) face uncertainty regarding compliance standards, making 2026 a pivotal year for clarifying which applications are economically viable now versus those requiring longer adoption timelines. Success will hinge on companies focusing on core market drivers such as cost, quality, performance, and customer experience rather than merely reacting to shifting regulations. A key trend is the increasing emphasis on American manufacturing as a competitive advantage, driven by domestic content requirements tied to federal infrastructure funding programs like Build America Buy America (BABA). These mandates, with waivers expiring or under review, require fleet operators to source from suppliers with verified domestic manufacturing and supply chains. This shift is expected to intensify, favoring manufacturers with established U.S.-based operations and compliance documentation. Additionally, the rise of AI data centers is creating new demand for commercial battery technology, prompting battery manufacturers to repurpose or expand U.S
energyelectrificationbattery-technologycommercial-electrificationdomestic-manufacturingAI-data-centersenergy-storageTrump Media pivots to fusion energy with $6B AI-focused merger
Trump Media & Technology Group is making a significant strategic shift by merging with fusion energy startup TAE Technologies in a deal valued at over $6 billion. This all-stock merger will create one of the first publicly traded nuclear fusion companies, combining Trump Media’s social media platform Truth Social with TAE’s experimental clean energy technology. The new entity plans to build a utility-scale fusion power plant as early as next year, aiming to supply electricity to AI data centers, which have rapidly growing energy demands. TAE brings substantial scientific expertise and financial backing from major investors like Google, Chevron, and Goldman Sachs, having already developed five fusion reactors and raised over $1.3 billion in private funding. The merger has sparked concerns about conflicts of interest given the regulatory and public funding requirements of fusion energy, especially with Donald Trump owning about 41% of Trump Media stock. Critics warn that Trump’s involvement could complicate government oversight. The announcement coincides with regulatory moves to facilitate direct connections between tech companies’ data centers
energynuclear-fusionAI-data-centersclean-energyfusion-reactorsenergy-mergerTAE-TechnologiesBoom Supersonic launches 42MW turbine for AI data center power surge
Boom Supersonic, traditionally known for its supersonic passenger aircraft development, has expanded into the energy sector by launching a new 42 MW natural gas turbine called Superpower. This turbine is designed specifically to meet the surging electricity demands of AI data centers, which are increasingly constrained by regional power supply limitations. The Superpower turbine leverages the high-temperature core technology from Boom’s Symphony supersonic engine, enabling it to maintain full power output even in extreme heat (e.g., 110°F) without the need for water cooling—a significant improvement over legacy turbines that lose 20-30% efficiency under such conditions. The company has secured a $1.25 billion order from Crusoe, a major AI data center operator, for 29 Superpower turbines totaling over 1.21 gigawatts of planned power capacity, with deliveries expected to begin in 2027. Boom plans to scale production rapidly, aiming for two gigawatts of annual output initially and targeting four gigawat
energygas-turbineAI-data-centerspower-generationBoom-Supersonicnatural-gasindustrial-turbineWorld’s first fast-neutron nuclear reactor to power AI data centers
French startup Stellaria has secured its first power reservation from Equinix for Stellarium, the world’s first fast-neutron molten-salt nuclear reactor designed to reduce nuclear waste. Stellarium is a fourth-generation reactor using liquid chloride salt fuel in a closed fuel cycle, capable of destroying more long-lived nuclear waste than it produces—a milestone no commercial reactor has yet achieved. The compact reactor, occupying only four cubic meters, can utilize a wide range of nuclear fuels including uranium, plutonium, MOX, minor actinides, and thorium. Stellaria aims to achieve its first fission reaction by 2029 and begin commercial deployment by 2035. Key design features include passive cooling, four physical containment barriers, and operation for over 20 years without refueling, making it a safe, resilient, and scalable clean energy source. The agreement with Equinix will enable the company’s AI data centers to operate autonomously with carbon-free, controllable nuclear power, supporting sustainable and dec
energynuclear-reactorfast-neutron-reactorclean-energyAI-data-centerssustainable-energymolten-salt-reactorMicro nuclear reactor targets 100-MW power for AI data centers
Terra Innovatum, a developer of micro-modular nuclear reactors, and Uvation, a global AI infrastructure technology provider, have signed a Letter of Intent to launch a 1 MWe pilot program using Terra Innovatum’s SOLO micro-reactor to power AI data centers. This initial phase includes an option to scale up to 100 MWe to meet Uvation’s growing data center power needs. The SOLO reactor, developed over six years with a focus on safety and licensing, uses commercial off-the-shelf components to streamline deployment and supports evolving nuclear fuels such as LEU+ and HALEU. The modular design allows for combined installations capable of delivering power in the gigawatt range, addressing forecasts of AI infrastructure power demand exceeding 1 GW. The partnership aims to solve the critical power shortage faced by AI data centers, which traditional electrical grids struggle to support due to the high-density, high-performance computing requirements. By adopting a “behind-the-meter” strategy—installing reactors directly at data
energymicro-nuclear-reactorAI-data-centersmodular-powerenergy-infrastructurenuclear-energypower-supplyHow much of the AI data center boom will be powered by renewable energy?
The article discusses the rapid growth of AI data centers and their significant power demands, highlighting that global spending on data centers is projected to reach $580 billion this year—surpassing investments in new oil exploration by $40 billion. This shift underscores the evolving global economy and raises concerns about the strain on electrical grids, especially as many data centers are planned near large urban populations. The majority of electricity demand from these centers is expected to come from the U.S., with China and Europe also contributing substantially. The article emphasizes that renewable energy adoption will likely be driven more by business considerations than environmental policies. A key potential positive noted is the opportunity for innovation in renewable energy and data center design, with companies like Redwood Materials launching initiatives such as Redwood Energy to create microgrids powered by repurposed EV batteries specifically for AI data centers. This approach could help mitigate grid stress, particularly in regions prone to power shortages like Texas. However, questions remain about how many planned data centers will actually be built given the enormous
energyrenewable-energydata-centersAI-data-centerselectrical-gridsmicrogridsbattery-recyclingElon Musk has new plans for Earth orbit. Astronomers are already scared
Elon Musk has proposed two ambitious new uses for SpaceX’s Starlink satellite network that aim to address pressing global challenges: the soaring energy demand of AI data centers and climate change mitigation. Musk envisions scaling up Starlink’s advanced V3 satellites with high-speed laser links to effectively relocate data centers into orbit, thereby reducing the unsustainable power consumption on Earth driven by AI technologies. Additionally, he suggested deploying an AI-powered satellite constellation to cool the planet and combat global warming through solar geoengineering techniques. These proposals align with Musk’s pattern of targeting large-scale problems through his companies but raise significant technical, environmental, and ethical concerns. While the idea of space-based data centers has been floated by other tech leaders, including former Google CEO Eric Schmidt, experts remain skeptical about its practicality. Astronomers point out the logistical challenges of maintaining and repairing hardware in orbit, as well as the risks posed by space debris and the environmental impact of launching vast amounts of material into space. Although some research teams argue
energysatellite-technologyAI-data-centersSpaceXglobal-warmingspace-infrastructuresustainable-energy-solutionsEven after Stargate, Oracle, Nvidia and AMD, OpenAI has more big deals coming soon, Sam Altman says
OpenAI has been actively securing large-scale infrastructure deals to support its rapidly growing AI model development, with major partnerships involving Nvidia, AMD, Oracle, and others. Nvidia has invested in OpenAI, becoming a shareholder, while AMD has granted OpenAI up to 10% of its stock in exchange for collaboration on next-generation AI GPUs. These deals include commitments for tens of gigawatts of AI data center capacity, such as OpenAI’s $500 billion Stargate deal with Oracle and SoftBank for U.S. facilities, and additional expansions in the UK and Europe. Nvidia is also preparing OpenAI for a future where it operates its own data centers, although the cost of such infrastructure—estimated at $50 to $60 billion per gigawatt—is currently beyond OpenAI’s direct financial capacity. OpenAI CEO Sam Altman emphasized that these partnerships are part of an aggressive infrastructure investment strategy to support more capable future AI models and products. Despite OpenAI’s revenue not yet approaching the scale of its
energyAI-data-centersNvidiaAMDOpenAIcloud-computingsemiconductor-chipsUS plans to power AI data centers with advanced nuclear energy
The US National Nuclear Security Administration (NNSA) has issued a Request for Proposals (RFP) to develop and operate large-scale Artificial Intelligence (AI) data centers powered by advanced nuclear energy. The initiative targets two historic nuclear sites—the Savannah River Site (SRS) in South Carolina and the Oak Ridge Reservation in Tennessee—with ten tracts of land at SRS identified for potential development. This effort aims to strengthen American leadership in AI and energy by fostering public-private partnerships that integrate innovative on-site energy generation and storage solutions, aligning with the Department of Energy’s goals for reliable, energy-efficient, and water-conscious operations to meet the substantial power demands of AI technologies. Selected private partners will be responsible for the full lifecycle of the projects, including construction, operation, and decommissioning, and must secure their own utility interconnections. Proposals will be competitively evaluated based on technological readiness, financial viability, and regulatory compliance. The Savannah River Site, historically a key facility for nuclear weapons material production
energynuclear-energyAI-data-centersadvanced-energy-infrastructureDepartment-of-Energynuclear-securitysustainable-power-generationUS nuclear firm to build 15 MW reactors a mile underground by 2026
Deep Fission, a US nuclear startup, has secured $30 million through a reverse merger with Surfside Acquisition Inc. to develop and deploy 15-megawatt mini nuclear reactors buried a mile underground. These reactors, designed to fit into 30-inch diameter boreholes, use pressurized water cooling and leverage proven technologies from nuclear submarines and traditional plants. By situating the reactors deep in bedrock, the company aims to enhance safety through natural shielding, reduce surface footprint, and protect against external threats. The design incorporates elements from nuclear, oil and gas, and geothermal industries, using off-the-shelf parts and low-enriched uranium to simplify supply chains. Deep Fission projects electricity costs of 5 to 7 cents per kilowatt-hour for its commercial systems. The company plans to complete its first pilot reactor by July 2026, supported by a Department of Energy (DOE) Reactor Pilot Program that aims to streamline permitting and accelerate deployment. Deep Fission has also signed a deal
energynuclear-energyunderground-reactorssmall-modular-reactorsclean-energyAI-data-centersDepartment-of-EnergyNuclear reactors to power AI data farms in US' first private HyperGrid
Fermi America, in partnership with South Korea’s Hyundai Engineering & Construction (Hyundai E&C) and the Texas Tech University System, is developing a landmark private energy campus near Amarillo, Texas, called the “HyperGrid.” This project aims to power next-generation AI data centers using a hybrid energy system that integrates America’s largest private nuclear power complex, the nation’s biggest combined-cycle natural-gas facility, solar arrays, battery storage, and grid power. The campus is designed to deliver one gigawatt of power by the end of 2026, primarily consumed on-site by AI data centers rather than fed into the public grid. Central to the HyperGrid are four Westinghouse AP1000 pressurized-water nuclear reactors, for which the U.S. Nuclear Regulatory Commission has accepted Fermi America’s Combined Operating License Application, marking a record-speed review. Construction of the nuclear island is planned to start in 2026, with the first reactor expected to be operational by 2032.
energynuclear-powerAI-data-centersHyperGridrenewable-energybattery-storagecombined-cycle-natural-gasNuclear-powered hydrogen explored to fuel global clean energy shift
The article discusses First Hydrogen Corp.'s new initiative to design small modular nuclear reactors (SMRs) in collaboration with the University of Alberta, aiming to produce low-carbon “green” hydrogen at scale. This partnership focuses on optimizing SMR technology—compact, factory-built nuclear reactors producing up to a few hundred megawatts—to generate the heat and electricity needed for hydrogen production without carbon emissions. The project targets cost-competitive hydrogen generation to support growing energy demands, particularly from artificial intelligence (AI) data centers, which Goldman Sachs predicts will increase power consumption by 160% by 2030 and could account for up to 4% of global electricity use. SMRs offer advantages such as modular construction, reduced accident risks, longer fuel cycles, and suitability for locations where large reactors are impractical. Canada, with its 60-year nuclear safety record and government backing, is positioning SMRs as a key element of future energy independence. Several provinces are advancing SMR projects, and Prime Minister Mark Car
energynuclear-powerhydrogen-productionsmall-modular-reactorsclean-energyAI-data-centerssustainable-energyNuclear-powered hydrogen explored to fuel global clean energy shift
Canada-based First Hydrogen Corp. has initiated a collaboration with the University of Alberta to design small modular nuclear reactors (SMRs) aimed at producing low-carbon "green" hydrogen at competitive costs. This partnership focuses on refining reactor fuels, core materials, and plant layouts to optimize SMRs for converting heat and electricity into hydrogen without carbon emissions. First Hydrogen’s move into nuclear technology, through its newly formed subsidiary First Nuclear, targets large-scale hydrogen production to meet the growing electricity demands of AI-driven data centers, which Goldman Sachs predicts will increase data center power consumption by 160% by 2030, potentially accounting for up to 4% of global electricity use. SMRs, producing up to a few hundred megawatts, offer advantages over traditional gigawatt-scale reactors due to their modular, factory-built design that allows easier onsite assembly and installation in diverse locations such as industrial campuses or remote areas. Their simplified, mostly underground construction aims to enhance safety, reduce refueling frequency, and lower upfront costs
energynuclear-energyhydrogen-productionsmall-modular-reactorsclean-energyrenewable-energyAI-data-centersGM Takes Aim At Tesla's Energy Storage Business, Too
General Motors (GM) is strategically expanding into the utility-scale energy storage market, directly challenging Tesla’s Megapack business. This move comes amid Tesla’s recent struggles in electric vehicle (EV) sales, exemplified by GM’s GMC electric Hummer reportedly outselling Tesla’s Cybertruck in Q2. GM’s new collaboration with Redwood Materials, founded by Tesla co-founder and former CTO JB Straubel, aims to develop large-scale energy storage systems tailored to meet the surging power demands of AI data centers and other applications. This partnership leverages both newly manufactured GM batteries and second-life EV batteries, positioning GM to capitalize on growing grid-scale battery needs amid uncertain EV market conditions and potential declines in EV tax credits. The collaboration builds on prior efforts where GM and Redwood deployed second-life EV batteries in Nevada to power AI infrastructure, creating the largest second-life battery system globally and North America’s biggest microgrid. Redwood’s business model focuses on lithium-ion battery recycling and repurposing, supported by significant funding
energyenergy-storageelectric-vehiclesbatteriesGMTeslaAI-data-centersUS firms to give used EV batteries a second life to feed AI hunger
General Motors (GM) and Redwood Materials have partnered to repurpose used electric vehicle (EV) batteries into large-scale energy storage systems, addressing a critical environmental challenge posed by the upcoming influx of first-generation EV batteries. This initiative supports GM’s sustainability strategy by providing a second life for retired automotive battery packs, which typically retain substantial energy capacity even after falling below optimal levels for vehicle use. The repurposed batteries are integrated into microgrids—localized, independent power grids that offer highly reliable, uninterrupted electricity—crucial for energy-intensive operations such as AI data centers. A notable project in Sparks, Nevada, already powers a 63 megawatt-hour microgrid supplying energy to an AI infrastructure company, marking the largest such installation in North America. The program tackles the growing demand for energy storage solutions driven by the rapid expansion of AI data centers, which are expected to triple their share of U.S. electricity consumption from 4.4% in 2023 to 12% by
energyelectric-vehicle-batteriesenergy-storagemicrogridssustainabilitybattery-recyclingAI-data-centersHair-thin chip transfers 100 million books in 7 minutes with just a spark of power
Scientists at Canada’s Université Laval have developed a groundbreaking optical chip that can transmit data at an unprecedented speed of 1,000 gigabits per second (Gbps) while consuming minimal energy—just four joules, comparable to heating one milliliter of water by one degree Celsius. This hair-thin chip uses pairs of microring modulators made from silicon to manipulate both the intensity and phase of light, enabling a dual-channel approach that vastly increases bandwidth within a compact size. This technology represents a significant leap from current systems, which typically max out at around 56 Gbps, allowing the transfer of data equivalent to 100 million books in under seven minutes. The innovation addresses the growing energy demands of AI data centers, where thousands of processors must communicate over long distances, resulting in extensive infrastructure and high power consumption. By enabling faster and more efficient communication as if processors were only meters apart, the chip could dramatically reduce energy use and physical footprint in AI systems. While still in the laboratory
energyphotonic-chipoptical-communicationdata-transmissionenergy-efficiencysilicon-microring-modulatorsAI-data-centersUS to launch world’s largest power project using nuclear, solar, gas
Fermi America, a Texas-based energy company co-founded by former U.S. Energy Secretary and Texas Governor Rick Perry, has announced plans to build the world’s largest energy and data campus, called the “Hypergrid,” near the DOE’s Pantex nuclear weapons plant in Amarillo, Texas. Covering 5,800 acres, the facility will integrate multiple energy sources—nuclear, natural gas, solar, and wind—to power 18 million square feet of AI data centers with up to 11 gigawatts of IT capacity, sufficient to supply electricity to over 8.2 million homes. The project is launched in partnership with the Texas Tech University System and aims to address the growing U.S. demand for AI infrastructure while enhancing national energy security. The Hypergrid project is positioned as a strategic response to global energy competition, particularly highlighting the U.S. lagging behind China in nuclear reactor development. The site’s proximity to the Pantex DOE facility and major natural gas fields underscores
energynuclear-powersolar-energynatural-gasAI-data-centersenergy-infrastructurerenewable-energyRedwood Materials launches energy storage business and its first target is AI data centers
Redwood Materials, founded by former Tesla CTO JB Straubel, has launched a new energy storage business called Redwood Energy, targeting AI data centers as its initial customers. The company is repurposing thousands of retired EV batteries—currently stockpiling over 1 gigawatt-hour and expecting an additional 4 gigawatt-hours soon—to create large-scale, clean energy storage systems. Their first project, in partnership with AI infrastructure firm Crusoe, involves a 12 MW, 63 MWh microgrid in Nevada that powers a modular data center using energy stored from an adjacent solar array. This operation is already profitable and marks a significant expansion beyond Redwood’s core battery recycling and materials supply business. Redwood Materials has built a circular supply chain by recycling battery scrap and consumer electronics to extract valuable materials like cobalt, nickel, and lithium, which it then sells to major manufacturers including Panasonic, Toyota, and GM. The company has also moved into cathode production and expanded its footprint globally. Redwood Energy
energyenergy-storageEV-batteriesbattery-recyclingmicrogridAI-data-centersrenewable-energyTrump signs executive orders to revive US’ nuclear power leadership
energynuclear-powerreactor-technologyAI-data-centersuranium-miningregulatory-reformrenewable-energy