AI Power Demand Crisis Forces Data Centers to Buy Ship Engines for Off-Grid Solutions

AI is so power-hungry it's buying ship engines now

The AI power crunch has a new symptom, and it comes from a shipyard. A South Korean shipbuilder just sold $673.8 million of engines to run one US tech company's data centres off the grid. When the utility cannot connect you for years, you buy roughly a hundred gas engines and make your own gigawatt. HD Hyundai Heavy Industries, the shipbuilding arm of HD Hyundai, announced the contract on 9 August. The buyer is Corban Energy Group, a New Jersey developer that supplies gas, LNG and power gear to data centres and defence projects. The kit will feed data centres run by an unnamed "major U.S. technology company." The numbers are the story. The order covers 1,000 megawatts of capacity, built around HD Hyundai's 9.6-MW HiMSEN engine. That points to something like a hundred engines. It is the largest power-generation engine contract in the company's history, and its second US data-centre deal in four months. Prime power, not backup This is not a diesel generator waiting out a blackout. A 1,000 MW order of medium-speed engines is prime power, meant to run a campus around the clock. The engines start fast, follow load, and scale in roughly 10 MW blocks. That suits a site that cannot pause and cannot wait for the grid. That last point is the whole reason the deal exists. Grid interconnection queues now stretch into years, so operators contract for their own generation instead. It is the same logic behind Amazon's off-grid gas plant in Texas, and behind the state grid audits now slowing approvals. HD Hyundai's first US move came in April: about 684 MW for $425 million, with developer Aperion Energy Group, per Unite.AI. The Corban deal is bigger on every axis. Two record orders in four months turn a one-off into a strategy. The catch is the fuel These are natural-gas engines. So a campus powered this way ties its running costs and its emissions to gas. That is exactly the trade-off driving the local backlash against data centres, where power bills and pollution top the list of complaints. Demand is what makes the gamble look safe. The Electric Power Research Institute projects data centres rising from 4-5% of US electricity today to 9-17% by 2030. HD Hyundai is spreading across that curve. Its shipbuilding holding company is developing floating data centres, while other affiliates move into power distribution and engine servicing. The through-line is simple. When a tech company needs a gigawatt and the grid cannot promise a date, buying a hundred engines stops being an edge case. It becomes the plan. The faster that plan scales, the more of AI's power problem gets answered with gas.

AI's power crunch is sending data-centre builders to the banks

Electricity, not chips, is now the binding constraint on AI, and the scramble to secure it is driving builders to seek billions in financing pledges. The bottleneck in AI is no longer chips. It is power, and the scramble to secure it is sending data-centre builders to the banks for billions in financing commitments, Bloomberg reports.The shift is a big one. For much of the AI boom the scarce resource was Nvidia's GPUs; now it is electricity, along with the contracts, substations, and financing needed to deliver it at scale. The numbers are staggering. The four largest cloud firms are set to spend around $725bn on AI infrastructure this year alone, and single sites now carry price tags that would once have funded a whole company, like Meta's $13bn Texas data centre. Power is what makes those sites so expensive. Server racks that drew about 3kW for ordinary computing now pull up to 150kW for AI, and global data-centre capacity is expected to nearly double to roughly 200GW by 2030. That is where the banks come in. Builders want firm financing pledges before they commit to power deals and construction contracts, and the sums are large enough that no single balance sheet wants to carry them alone. The borrowing has already reshaped credit markets. Lenders have poured into the sector, from a $5.9bn loan for one data-centre operator to a wave of bonds and private-credit deals feeding the build. The financing is getting creative, and opaque. Oracle's $16.3bn data-centre raise leaned on private credit after banks grew cautious, a sign of how far off the traditional path the money now travels. Much of it sits out of plain view. Analysts estimate the sector is carrying around $1.65tn in off-balance-sheet obligations, structured through special vehicles that keep the debt off the tech giants' own accounts. The hunt for power is reshaping energy markets. Builders are reviving gas plants, signing nuclear deals, and jumping utility queues, turning data centres into some of the largest new electricity buyers in a generation. The strain is landing on everyone else. Grids are creaking under the new load, and in some regions households are already paying more as data centres compete for the same electrons. Regulators are watching uneasily. The Bank for International Settlements has flagged 'circular financing' between cloud firms, their suppliers, and construction lessors as one of the bigger risks to financial stability. The building itself is a chokepoint. Fewer than ten firms in the world can deliver a hyperscale project, and one of them, Turner, is sitting on a record backlog of about $44bn with a large share tied to data centres. Banks are not lending blindly, either. As the sums swell and the structures grow more circular, some lenders have pulled back, which is part of why builders are chasing firm pledges rather than assuming the money will be there. Power is being locked up years ahead. Builders are signing long deals for electricity, gas, and even nuclear output, and the bank pledges are what make those commitments credible. Underneath it all is a bet that demand shows up. The spending only pays off if AI usage keeps climbing fast enough to fill the capacity, which is why every quarter of AI revenue is now watched so closely. There is a self-reinforcing quality to it, too. The more money commits to the build, the more the industry needs AI to keep growing to justify it, which is precisely what makes the circularity regulators fear so hard to unwind. The scale is historic. Goldman Sachs sees cumulative spending on this buildout running into the trillions by 2030, a figure that dwarfs past infrastructure booms and leaves little room for error. For now, the money keeps flowing toward the wall socket. The race for compute has become a race for power, and the winners may be decided less by who has the best models than by who can finance the electricity to run them.

Enabling the next generation of AI data centers

Exploring the infrastructure trade-offs behind AI data center growth Artificial intelligence (AI) is reshaping the scale and complexity of data center infrastructure. Traditional data center facilities were designed around relatively steady CPU workloads and predictable growth in power demand, allowing developers to secure energy supply alongside growing demand, cooling systems based on known and mature technology and infrastructure capacity with a reasonable degree of certainty. AI workloads, however, demand far more power with greater energy density. Electricity consumption from data centers has grown at 12% per year over the last five years and expected demand growth, particularly in AI training data centers, is set to drive a substantial increase in power demand. Meeting this demand, while maintaining efficiency, is pushing developers toward gigawatt-scale data centers and with the timeframe for delivering these facilities rapidly compressing, the ability to deliver new infrastructure efficiently is increasingly critical. In addition, as the scale of these developments grows, so does the complexity of delivering them. Grid interconnections can delay timelines by years, equipment supply chains are stretched, and projects must meet stringent reliability targets while navigating regulatory, environmental and community requirements that vary by region and country. For owners and developers, the challenge is no longer simply constructing another data center building. The next generation of AI data centers requires a fully integrated approach across power, cooling, transmission, water, digital systems and long-term operations. Success depends on designing these facilities as resilient, flexible and energy-optimized industrial campuses. Balancing site trade-offs to unlock faster delivery Site selection is one of the clearest expressions of this dynamic where teams are typically assessing a series of imperfect options, each with its own advantages and constraints. For example, one site may offer lower cost land but lack the existing infrastructure required to support large scale development, while another may provide access to grid power but at a significantly higher cost or with timelines that delay delivery. In practice, few locations offer everything required, and selecting a site becomes an exercise in understanding what should be prioritized, what can be mitigated, and what must be accepted. Factors like water availability, land constraints, fiber connectivity, permitting timelines and social license to operate are all deeply important to success. Developers must consider how to optimize within these constraints. Where grid power is unavailable or delayed, for example, off grid or hybrid energy solutions may be introduced. While these approaches can accelerate delivery, they also bring different capital requirements, financing structures, and operational considerations that must be carefully weighed. Combining power solutions can accelerate bringing capacity online more efficiently As AI workloads drive unprecedented levels of demand, power strategies also require a reassessment against expected scale timelines. Grid supply does offer lower long-term energy costs, stability and resilience advantages eventually but hinges on capacity constraints, and extended interconnection timelines. In contrast, behind the meter generation, such as gas turbines or reciprocating engines, can be deployed more quickly and provide greater operational control. This, however, comes with higher upfront capital requirements, higher operational costs, fuel dependencies and more complex permitting considerations. As speed-to-market is a key competitive driver, many large-scale developments are willing to pay a premium for off-grid or hybrid architectures, including battery storage and integration of renewables where accessible. These systems are coordinated through microgrid controls, allowing operators to manage load variability, maintain resilience through islanding, and optimize overall system performance. The final configuration is shaped by how factors such as time to market, grid availability, resilience, and overall cost evolve. Rethinking cooling can support high-density AI and optimize when energy is used With this increase in power demands comes a corresponding increase in heat generation. The physics and economics of air cooling are struggling to keep pace with the thermal loads generated by AI workloads, forcing a shift toward alternative solutions. One solution is liquid cooling, which is gaining traction as a more effective way to manage higher heat loads. Transferring heat more efficiently, it enables facilities to operate at the densities required by AI infrastructure. However, it does also introduce new dependencies, particularly around liquid cooling solutions and the infrastructure required to support it. At the same time, taking a broader view of cooling opens up new opportunities. Cooling systems can be integrated with wider power infrastructure, excess heat can be connected to industrial processes that can utilize it and waste heat from data centers can be repurposed for applications such as district heating, which is already quite common in the Nordics. Approaching cooling in this way allows developers to design systems that make better use of energy and create additional value through heat reuse and integration with surrounding infrastructure. Additionally, thermal energy storage gives AI data centers the ability to shift cooling demand away from peak periods by producing chilled water when electricity is cheaper or more available and using it later when loads are highest. This creates valuable demand response capability, allowing the facility to reduce its grid draw during periods of system stress, lower demand charges and support utility programs without impacting data center operations. In combination with batteries and advanced controls, thermal storage can help stabilize both the data center and the surrounding grid. Early efforts on permitting can identify the fastest development route and avoid delays Permitting and regulatory considerations sit alongside these technical decisions, shaping what is possible and how quickly projects can move forward. Requirements vary by region, country and project type, but in all cases, they influence how projects must be designed from the outset. For example, grid connected developments may be constrained by connection approvals and capacity limits, while sites incorporating on-site generation may require air quality or emissions permits that influence technology choices. Land use restrictions, environmental approvals and community considerations can further shape site layout, development timelines and even overall project viability. Addressing these requirements early, and in parallel with technical and commercial decision making, is therefore as important as those other factors. When permitting is treated as part of the initial planning process, it allows projects to be structured in a way that is both deliverable and aligned with regulatory expectations from the beginning. This, in turn, reinforces the need for a coordinated approach across the full range of stakeholders involved. Energy providers, technology companies, developers, regulators and local communities each play a role in shaping outcomes, and the interaction between them becomes a critical factor in how effectively projects can progress. Having the right expertise in place to connect these elements enables developers to navigate this complexity more effectively, ensuring that decisions made early on are aligned across disciplines. This early alignment helps create a more integrated delivery pathway, reducing friction between project phases and supporting smoother progression from planning through to construction and execution. Turning AI demand into operational capacity at the speed and scale the market requires The importance of this becomes clearer when looking at how these challenges play out in practice. In Texas, for example, early engineering work on a gigawatt scale AI training data center helped define the infrastructure strategy for one of the largest behind the meter energy systems supporting AI workloads. The project includes 5 GW of gas generation capacity, up to 1.25 GW of solar PV, utility scale battery storage and a microgrid supporting 20 buildings totaling 10 million square feet, each designed for around 250 MW of power demand. Projects of this scale reflect the sheer pace and ambition of AI demand, but ultimately, success comes down to how effectively that demand is translated into deliverable infrastructure. That means making early decisions that can withstand real world constraints, from power availability and permitting through to long term operational performance. Bringing these elements together into a coherent strategy, and aligning the stakeholders needed to deliver it, is what will enable projects to move at the speed and scale the market now requires. We've featured the best database software. This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

AI's volatile power demand is damaging its own data centers | Fortune

Rapid swings in AI data centers' power demands are straining vital equipment, causing batteries, generators and cooling systems to malfunction or wear out far sooner than expected. As the AI boom accelerates, these technical problems suggest added costs and unforeseen reliability problems, with even a few minutes of lost uptime hitting data-center developers' revenue. They come at a time when investors and lenders are already jittery about hyperscalers' hundreds of billions of dollars of spending, amid growing concerns that these facilities could be depreciating much faster than estimated. The problems are also a potential source of wider instability in power grids that are already straining to keep the lights on. "AI does create very unusual power demand," said Amber Villegas-Williamson, principal consultant at the Uptime Institute in the UK, which advises electricity suppliers and data centers on standards and reliability. "It's like over-revving your car wears out the engine faster than keeping a constant speed." Data centers have been around for decades, gulping down electricity while they ensure that everything from your favorite streaming show to your online grocery order functions smoothly. But facilities designed for AI computing are different because their demand is so large and swings much more dramatically. Power increments equivalent to the consumption of factories, towns or even cities can appear and disappear within seconds, creating repeated shocks that connected equipment struggles to absorb. A gigawatt data center is equivalent to a city the size of Boston, half of which can flicker on and off every few seconds, said Shannon Miller, founder and president of Mainspring Energy Inc., which works on micro-grid projects for industrial and data-center customers. Some AI campuses planned in Texas, the Midwest and other states are more than five times bigger, consuming nearly as much power on average as New York City. AI data centers put particular strain on their power supply when they are training new models -- a process that mobilizes all of the graphics processing units in unison. Like the digital equivalent of bees swarming or a school of fish changing direction, hundreds of thousands of GPUs can power up and down on a millisecond basis. AI at times sees power usage spike as much as 50% above its design capacity, "so a 1 gigawatt facility may use 1.5 gigawatts for a split second," said Drew Baglino, a former Tesla Inc. executive who started Heron Power Electronics Co. The company is developing equipment to manage power fluctuations for Nvidia Corp.'s even more energy intensive next-generation of servers, due in 2027. Most equipment isn't designed for such big swings in power consumption. Jon Parrella, chief executive officer of energy-storage developer Terraflow Energy, likens it to driving a Ferrari and shifting straight from sixth gear to first. "You can't swing that fast," he said. This story is based on interviews with more than three dozen power experts in the US and Europe, including generators and other power suppliers, data-center developers, grid operators, utilities, investors, standards developers, insurers and regulators, almost all of whom said the physical stresses on the facilities were evident. Cranks on small natural gas combustion engines used to generate power at data centers have broken off, several of those people said. At xAI's Colossus computing facility in Memphis, Tennessee, gas-fired turbines had developed cracks, one of the people said. Batteries were installed within the system to help smooth out power swings and reduce the strain on spinning turbines, the person said. SpaceX, the parent company of xAI, didn't respond to requests for comment. Turbines have also cracked at much smaller data centers in the UK, said Andrew Cunningham, CEO of GeoPura Ltd., which is providing hydrogen for use in fuel cells that smooth out power flows at some sites. Cracks or wear on devices can cause electrical arc flashes -- when a current jumps between conductors -- potentially damaging AI chips, said Jennifer Scanlon, CEO of UL Solutions Inc., which tests and certifies new technology. There is a suite of equipment such as batteries, capacitors, transformers and flywheels that can help stabilize power flows. However, in the bid to build AI computing capacity quickly not enough of these technologies are being used at new data centers, several of the people said. Batteries that have been installed for this purpose have sometimes needed to replaced within months or even weeks due to the high strain, according to the Uptime Institute and other people working with operators. This problem is seen at data centers around the world, from the Middle East and Africa, to Europe and the US, said Villegas-Williamson of the Uptime Institute. Reliability Problems These issues are already causing delays or curtailing operations - and therefore revenue - at some AI computing facilities. To ensure that a planned 2.67-gigawatt AI campus in West Texas can achieve the 99.999% reliability required by Microsoft Corp., extra time was baked into the schedule for engineering, said Chris James, CEO of Joulent Inc., which is developing the facility with energy giant Chevron Corp. This means power delivery will begin in 2028 instead of 2027, he said. If essential equipment breaks down prematurely, "the financial consequence is not primarily replacing a pump or a breaker or some power component -- it's the the value of that expensive compute capacity not generating revenue because it's offline," said Jason Hoffman, chief strategy officer at data-center builder and operator Switch. The cost of downtime in terms of lost revenue varies widely, with estimates ranging from thousands to hundreds of thousands of dollars per minute, depending on the type of facility and its workload. Data centers are built on the assumption that, once they are online, they will operate around the clock 365 days a year, said a person involved in the financing of such facilities. In reality, some are seeing uptime closer to 80%, and unless resolved this could hit investors in certain projects in the next 12 to 24 months, the person said. Any issues with reliability add to wider concerns about the returns generated from hundreds of billions of dollars of planned AI investments. The rate of depreciation of another crucial piece of equipment at data centers, the GPU racks themselves, has raised questions about whether the industry can be as profitable as it promises. These problems with reliability also have the potential to destabilize the wider power grid. This sprawling web of high voltage lines, transformers and power plants requires constant calibration, something that has become more challenging with each passing year due to aging equipment, rising demand and extreme weather. The expansion of intermittent wind and solar generation, which often result in big swings in supply from one hour to the next, are already a destabilizing force. AI data centers can put this volatility on steroids. "These loads are extremely dynamic or fluctuating, which causes grid instability and can lead to, if not corrected, potential blackouts or power outages," said Sreemant Roy, a power-quality expert and global offer manager at Schneider Electric in Nashville, Tennessee. Of particular concern is a data center's ability to cause sub-synchronous oscillations in the power flow, which can damage equipment connected to other parts of the network, he said. "That has made utility companies globally very worried," Roy said. Within the last two years, the North American Electric Reliability Corp. -- the top US regulatory body establishing standards aimed to keep the lights on -- has repeatedly warned and issued alerts that data centers are one of the greatest risks to grid stability. NERC evaluated more than 33 gigawatts of operational data centers in the US an found about three quarters of their load models "are insufficient to represent data-center dynamic behavior," according to a September report. Earlier this year, the agency issued a rare level-three alert requiring big data centers to address these immediate risks and submit their responses by Aug. 3. From top to bottom, the AI industry is aware of these issues and actively working on solutions. Nvidia started working more closely with power experts when it developed Blackwell GPUs, which were first released in 2024 and have become pervasive in data centers, said Dion Harris, senior director of hyperscale infrastructure solutions at the chipmaker. "We're building the chips and processors" but also using them in the company's own data centers, Harris said. Nvidia is working to make its deployments a lot smoother "both on the data-center build out, design and engineering phase, as well as on the power delivery." Data-center users have deployed techniques to smooth out the power fluctuations of AI workloads by running side computations -- essentially dummy math that isn't part of the training process -- to keep GPUs operating steadily. However, this approach has been criticized for wasting electricity at a time when power demand is surging. Last year, the National Laboratory of the Rockies near Denver, Colorado set up a test-bed on behalf of the US Department of Energy to figure out how to integrate AI safely onto the grid, said Martha Symko-Davies, the NLR's program manager for the DOE's office of electricity. The site has GPUs and power generation on site that developers can use to figure out if their setup can handle the variability of AI. One power supplier said they will use the facility to test batteries, software and other equipment intended to smooth out oscillations between the data center and grid that can cause damage on either side. "We have the opportunity right now to get it right," said Symko-Davies. To contact the author of this story: Naureen S Malik in New York at [email protected]

The power problem behind the AI boom: Why data centers are racing to upgrade their electrical infrastructure

A GPU cluster can be ordered, shipped, and racked in well under a year. Energizing the building that holds it can take three. That gap sits behind almost every data center headline of the past two years. Capital is not the constraint. Silicon, for the larger buyers, has mostly stopped being one. What actually decides whether a site opens on schedule is electricity, plus everything standing between a substation and a server cabinet: transformers, switchgear, breakers, busway, and an interconnection agreement that has to be signed before any of it means anything. A single rack now pulls what a small office used to For roughly two decades the planning numbers were comfortable. A typical enterprise rack drew five to ten kilowatts, and a hall designed around 100 to 150 watts per square foot could absorb whatever tenants brought in. Facilities were built with headroom because headroom was cheap. AI erased that headroom in about one product cycle. Dense accelerator racks now land somewhere between 40 and 130 kW, and vendor roadmaps discuss cabinets measured in the hundreds of kilowatts as a near-term expectation rather than a thought experiment. Density is not only a cooling problem, though cooling is the part that gets photographed. Each step up in per-cabinet power pushes work back upstream, onto distribution equipment that was specified for a building with a different purpose. The waiting list that starts long before the utility says yes Operators tend to discover the equipment problem in the wrong order. The interconnection study gets attention first, because it has a formal process attached and a queue position that can be quoted to investors. Then procurement comes back with lead times. Large power transformers are the usual shock. Buyers who were quoting under a year before 2020 have spent recent cycles quoting in multiples of that, and the constraint is not simple to unwind: grain-oriented electrical steel, bushings, tap changers, skilled winding labor, and factory floor space that cannot be conjured in a quarter. For a voltage transformer manufacturer, expanding production is far more complex than simply adding another assembly line. Medium- and high-voltage transformers require specialized winding equipment, experienced personnel, lengthy qualification processes, and years of manufacturing know-how. That makes production capacity slow to expand even when demand surges, which is one reason lead times across the industry have stretched so dramatically. Utilities running long-term replacement programs often secure manufacturing capacity years in advance, leaving data center developers competing for available production slots rather than for transformers themselves.. Switchgear, breakers, generators, and even basic busway have stretched in sympathy. So developers started doing something that would have looked reckless in 2019: ordering transformers speculatively, against sites that are not fully designed, and warehousing them until a layout catches up. It is an ugly use of capital. It also works, and it has quietly become normal practice among the larger builders. AI training loads behave less like a data center and more like a steel mill Classic data center load is boring in the best way. Thousands of uncorrelated workloads average out, and the facility draws a fairly flat curve that utilities find easy to plan around. Training clusters can break that assumption. Tens of thousands of accelerators working through synchronized steps may ramp together and, in some large deployments, swing a substantial share of site load within a very short window when a job stalls, checkpoints, or fails outright. How pronounced this gets depends heavily on the scheduler, the framework, and how the operator has configured power management, so the behavior varies more between sites than the headlines suggest. Grid operators in several markets have flagged it regardless, and interconnection agreements increasingly carry ramp-rate limits, ride-through obligations, and power quality conditions that would have been unusual for a data center a decade ago. For the electrical room, this changes what "correctly sized" means. Equipment gets specified against step loads and harmonic content rather than a steady thermal rating. Battery systems installed as a bridge to generator start are being asked to smooth load swings as a routine function, and some operators now shape workloads in software specifically so the electrical side sees a gentler curve. Utilities notice which campuses can promise predictable behavior. It is a far easier interconnection conversation than the alternative. Why copper, not silicon, sets the limit inside the rack Here is the arithmetic that quietly rewrote rack design. Power is voltage multiplied by current, so at a fixed voltage, more kilowatts mean proportionally more amps. A 100 kW cabinet fed at 415 V three-phase needs roughly 140 A coming in, which is manageable. The problem starts after the power supplies convert it down. On a traditional 54 V DC busbar inside the rack, that same 100 kW works out to something in the region of 1,850 A, and every one of those amps has to travel through copper, connectors, and contact surfaces that heat up in proportion to resistance. Copper does not scale gracefully. Doubling the current means roughly quadrupling the resistive losses, so the conductor has to get dramatically thicker to keep the heat under control, which makes it heavier, more expensive, and harder to route through a cabinet already crowded with liquid cooling manifolds. At megawatt-class racks the busbar stops being a component and starts being a structural element. Raising the voltage is the only lever that actually moves. At 800 V, the same 100 kW draws about 125 A instead of 1,850, which is why higher-voltage DC distribution went from a niche argument to a mainstream design direction in the space of two product generations. The physics did not change. The power density finally made the tradeoff obvious. Three routes around the interconnection queue Most large projects combine at least two of the following, and rarely rely on one. What the table cannot show is how often the choice gets made under duress, and the public record is full of examples. xAI's Memphis site is the clearest version of route two. Rather than wait for grid capacity, the company installed gas turbines on the property to run the cluster, and then spent months answering air permit questions from regulators and local groups. Fast, and expensive in ways that do not appear on the electrical drawings. Amazon took route three when it bought a data center campus adjacent to the Susquehanna nuclear plant in Pennsylvania, paying a reported several hundred million dollars for a site whose main asset was proximity to generation that already existed. Microsoft's twenty-year agreement with Constellation to restart the shuttered unit at Three Mile Island belongs to the same logic: buying output from generation nobody has to newly interconnect. The pattern across all three is worth noting. None of these deals were about electricity being expensive. They were about electricity being unavailable on a schedule that matched a hardware order. The retrofit question colocation tenants ask too late Existing halls are where the mismatch gets uncomfortable. A tenant with a modest allocation of GPUs will ask about adding a few high-density racks, and the answer often has nothing to do with the racks. The floor may not carry the weight of liquid-cooled cabinets. The cooling loop may not exist at all. The distribution may be sized for a per-cabinet average the new hardware clears in the first row. And then there is the failure mode nobody expects, which is that the protection equipment becomes wrong in the other direction. Feeding a denser hall usually means a bigger transformer with lower impedance, and lower impedance means more available short-circuit current at the fault. Switchgear that was properly rated when it was installed can end up under-rated for the fault current the new supply can deliver, which is not a performance issue but a safety one. The fix ranges from current-limiting reactors to splitting the bus to replacing the lineup entirely, and none of those are cheap surprises to find halfway through a retrofit. Retrofits that succeed usually take a zone approach, carving out a high-density pod with its own distribution rather than lifting the whole floor. That also matches how the hardware actually arrives, which is to say gradually, and rarely in the order originally planned. Partial upgrades are not elegant. They are frequently the only version that gets built. Where the race is actually being run For most of the last decade, the binding question was who could get the chips. That question has an answer now, and it turned out to be the easy one. The harder one is being settled in transformer factory order books, in interconnection queues, and in permit hearings held in counties most people could not find on a map, and it determines which of the announced gigawatts ever switch on. The next stretch of the AI race will be decided less by who designs the faster chip than by who can get power to it first.

AI's volatile power demand is damaging its own data centers

AI data centres' volatile power demand is damaging batteries, turbines and other equipment, raising downtime costs, delaying projects and creating wider grid-stability risks. Rapid swings in AI data centers' power demands are straining vital equipment, causing batteries, generators and cooling systems to malfunction or wear out far sooner than expected. As the AI boom accelerates, these technical problems suggest added costs and unforeseen reliability problems, with even a few minutes of lost uptime hitting data-center developers' revenue. They come at a time when investors and lenders are already jittery about hyperscalers' hundreds of billions of dollars of spending, amid growing concerns that these facilities could be depreciating much faster than estimated. ALSO READ | India's data centre capacity rises over four-fold to 1,575 MW as AI demand accelerates: Govt The problems are also a potential source of wider instability in power grids that are already straining to keep the lights on. "AI does create very unusual power demand," said Amber Villegas-Williamson, principal consultant at the Uptime Institute in the UK, which advises electricity suppliers and data centers on standards and reliability. "It's like over-revving your car wears out the engine faster than keeping a constant speed." Data centers have been around for decades, gulping down electricity while they ensure that everything from your favorite streaming show to your online grocery order functions smoothly. But facilities designed for AI computing are different because their demand is so large and swings much more dramatically. ALSO READ | Data centres, AI infra become IT's next growth bet Power increments equivalent to the consumption of factories, towns or even cities can appear and disappear within seconds, creating repeated shocks that connected equipment struggles to absorb. A gigawatt data center is equivalent to a city the size of Boston, half of which can flicker on and off every few seconds, said Shannon Miller, founder and president of Mainspring Energy Inc., which works on micro-grid projects for industrial and data-center customers. Some AI campuses planned in Texas, the Midwest and other states are more than five times bigger, consuming nearly as much power on average as New York City. AI data centers put particular strain on their power supply when they are training new models -- a process that mobilizes all of the graphics processing units in unison. Like the digital equivalent of bees swarming or a school of fish changing direction, hundreds of thousands of GPUs can power up and down on a millisecond basis. AI at times sees power usage spike as much as 50% above its design capacity, "so a 1 gigawatt facility may use 1.5 gigawatts for a split second," said Drew Baglino, a former Tesla Inc. executive who started Heron Power Electronics Co. The company is developing equipment to manage power fluctuations for Nvidia Corp.'s even more energy intensive next-generation of servers, due in 2027. Most equipment isn't designed for such big swings in power consumption. Jon Parrella, chief executive officer of energy-storage developer Terraflow Energy, likens it to driving a Ferrari and shifting straight from sixth gear to first. "You can't swing that fast," he said. This story is based on interviews with more than three dozen power experts in the US and Europe, including generators and other power suppliers, data-center developers, grid operators, utilities, investors, standards developers, insurers and regulators, almost all of whom said the physical stresses on the facilities were evident. Cranks on small natural gas combustion engines used to generate power at data centers have broken off, several of those people said. At xAI's Colossus computing facility in Memphis, Tennessee, gas-fired turbines had developed cracks, one of the people said. Batteries were installed within the system to help smooth out power swings and reduce the strain on spinning turbines, the person said. SpaceX, the parent company of xAI, didn't respond to requests for comment. Turbines have also cracked at much smaller data centers in the UK, said Andrew Cunningham, CEO of GeoPura Ltd., which is providing hydrogen for use in fuel cells that smooth out power flows at some sites. Cracks or wear on devices can cause electrical arc flashes -- when a current jumps between conductors -- potentially damaging AI chips, said Jennifer Scanlon, CEO of UL Solutions Inc., which tests and certifies new technology. There is a suite of equipment such as batteries, capacitors, transformers and flywheels that can help stabilize power flows. However, in the bid to build AI computing capacity quickly not enough of these technologies are being used at new data centers, several of the people said. Batteries that have been installed for this purpose have sometimes needed to replaced within months or even weeks due to the high strain, according to the Uptime Institute and other people working with operators. This problem is seen at data centers around the world, from the Middle East and Africa, to Europe and the US, said Villegas-Williamson of the Uptime Institute. Reliability Issues These issues are already causing delays or curtailing operations - and therefore revenue - at some AI computing facilities. Extra time was baked into the schedule for engineering at a planned 2.67-gigawatt AI campus in West Texas, said Chris James, CEO of Joulent Inc., which is developing the facility with energy giant Chevron Corp. This means power delivery will begin in 2028 instead of 2027, he said. "Data centers and the power supply cannot be built independently," he said in a subsequent statement. "The load, generation, storage, controls and grid connection all affect one another. As AI infrastructure scales, the projects that perform best will be the ones that account for those interactions early, rather than trying to solve them after construction." If essential equipment breaks down prematurely, "the financial consequence is not primarily replacing a pump or a breaker or some power component -- it's the the value of that expensive compute capacity not generating revenue because it's offline," said Jason Hoffman, chief strategy officer at data-center builder and operator Switch. The cost of downtime in terms of lost revenue varies widely, with estimates ranging from thousands to hundreds of thousands of dollars per minute, depending on the type of facility and its workload. Data centers are built on the assumption that, once they are online, they will operate around the clock 365 days a year, said a person involved in the financing of such facilities. In reality, some are seeing uptime closer to 80%, and unless resolved this could hit investors in certain projects in the next 12 to 24 months, the person said. Any issues with reliability add to wider concerns about the returns generated from hundreds of billions of dollars of planned AI investments. The rate of depreciation of another crucial piece of equipment at data centers, the GPU racks themselves, has raised questions about whether the industry can be as profitable as it promises. These problems with reliability also have the potential to destabilize the wider power grid. This sprawling web of high voltage lines, transformers and power plants requires constant calibration, something that has become more challenging with each passing year due to aging equipment, rising demand and extreme weather. The expansion of intermittent wind and solar generation, which often result in big swings in supply from one hour to the next, are already a destabilizing force. AI data centers can put this volatility on steroids. "These loads are extremely dynamic or fluctuating, which causes grid instability and can lead to, if not corrected, potential blackouts or power outages," said Sreemant Roy, a power-quality expert and global offer manager at Schneider Electric in Nashville, Tennessee. Of particular concern is a data center's ability to cause sub-synchronous oscillations in the power flow, which can damage equipment connected to other parts of the network, he said. "That has made utility companies globally very worried," Roy said. Within the last two years, the North American Electric Reliability Corp. -- the top US regulatory body establishing standards aimed to keep the lights on -- has repeatedly warned and issued alerts that data centers are one of the greatest risks to grid stability. NERC evaluated more than 33 gigawatts of operational data centers in the US an found about three quarters of their load models "are insufficient to represent data-center dynamic behavior," according to a September report. Earlier this year, the agency issued a rare level-three alert requiring big data centers to address these immediate risks and submit their responses by Aug. 3. From top to bottom, the AI industry is aware of these issues and actively working on solutions. Nvidia started working more closely with power experts when it developed Blackwell GPUs, which were first released in 2024 and have become pervasive in data centers, said Dion Harris, senior director of hyperscale infrastructure solutions at the chipmaker. "We're building the chips and processors" but also using them in the company's own data centers, Harris said. Nvidia is working to make its deployments a lot smoother "both on the data-center build out, design and engineering phase, as well as on the power delivery." Data-center users have deployed techniques to smooth out the power fluctuations of AI workloads by running side computations -- essentially dummy math that isn't part of the training process -- to keep GPUs operating steadily. However, this approach has been criticized for wasting electricity at a time when power demand is surging. Last year, the National Laboratory of the Rockies near Denver, Colorado set up a test-bed on behalf of the US Department of Energy to figure out how to integrate AI safely onto the grid, said Martha Symko-Davies, the NLR's program manager for the DOE's office of electricity. The site has GPUs and power generation on site that developers can use to figure out if their setup can handle the variability of AI. One power supplier said they will use the facility to test batteries, software and other equipment intended to smooth out oscillations between the data center and grid that can cause damage on either side.

AI's volatile power demand is damaging its own data centers

Artificial intelligence's tremendous hunger for electricity is already well known. What's less familiar is how the rapid fluctuations in data centers' appetites can break essential equipment at the facilities. Batteries, generators, cooling units and other critical systems are put under such strain at AI computing facilities that they are malfunctioning or prematurely reaching the end of their lives. As the AI boom accelerates, these technical problems suggest added costs and unforeseen reliability problems, with even a few minutes of lost uptime hitting data-center developers' revenue. They come at a time when investors and lenders are already jittery about hyperscalers' hundreds of billions of dollars of spending, amid growing concerns that these facilities could be depreciating much faster than estimated.

Data centers now face a power race: AI racks need 50 kW or more

Gigawatt campuses, grid delays and liquid cooling are redefining expansion Data center infrastructure has moved into a different phase. This isn't the kind of steady server growth the industry used to plan around. It calls for far more electricity, cooling, land, and supporting infrastructure than traditional expansion ever did. Older central processing unit (CPU)-focused facilities were usually built for 5 kW to 10 kW per rack. Racks loaded with accelerators now often need 50 kW or more, and some designs are already heading toward 100 kW. Power has become the main bottleneck. Data center electricity use has climbed about 12% a year over the past five years, could reach 565 TWh in 2026, and may go past 1,200 TWh by 2030. The server side is shifting even faster: accelerated server platforms alone are expected to jump 84% in 2026 and overtake conventional hardware by 2027. Those power and infrastructure requirements are pushing operators toward gigawatt-scale campuses. To make those sites work, they have to bring together power, water, cooling, fiber, permitting, and community support all at once. Grid interconnection delays and limited capacity are making that harder, so some operators are starting to look at on-site generation, off-grid designs, or hybrid energy. Cooling is changing just as fast. Air systems are running into their limits, and direct-to-chip or immersion cooling is becoming a core option for 100 kW-class racks. NVIDIA's latest infrastructure is reportedly fully liquid cooled. If you work in data center planning, this is a shift to keep close track of, especially with the market projected to grow from $147.3 billion in 2025 to $810.6 billion by 2033, while cost and sustainability pressure keep rising across global hyperscale and AI facilities.