AI's Power Hunger Hits Gas Turbine Shortage
· relationships
The Dark Side of AI’s Power Hunger
The tech industry’s obsession with artificial intelligence has been met with warnings about bias, job displacement, and ethics. However, another critical issue is emerging: the growing demand for massive amounts of electricity to power these behemoths.
Data centers are the backbone of the AI revolution, but they’re also a significant contributor to our energy crisis. Goldman Sachs recently highlighted the alarming rate at which data center power demand is increasing – from 31 gigawatts in 2025 to 66 GW by 2027. By then, these facilities will consume nearly 9% of total U.S. peak summer demand.
The problem lies not with the data centers themselves but with the equipment needed to generate this power: gas turbines. The “Big Three” manufacturers – GE Vernova, Siemens Energy, and Mitsubishi Heavy Industries – are booked solid until the 2030s. Their combined backlog is staggering: 220 GW of capacity commitments, but these numbers are not as straightforward as they seem.
GE Vernova has a firm equipment backlog of 53 GW and an additional 63 GW worth of slot reservations, which are essentially paid options that haven’t yet converted to orders. Siemens Energy’s 69 GW is a firm backlog with no reservations mixed in. Mitsubishi Heavy Industries’ 35 GW covers only large-frame turbines, leaving out its other product lines.
This discrepancy highlights the complexity and lack of transparency surrounding these numbers. It is clear, however, that the industry is facing a critical shortage of gas turbines. The Big Three are prioritizing high-end clients and “selective” projects, leaving smaller players and emerging markets in the lurch.
The implications are far-reaching: as data centers continue to proliferate, they will consume an increasingly large share of the global energy pie. If we can’t deliver the necessary infrastructure – or worse, prioritize high-margin contracts over timely delivery – we risk creating a perfect storm of inefficiency and waste.
This is not just about AI; it’s about our collective inability to manage growth. We’ve seen this story play out before: the dot-com bubble, the housing market collapse, and the current era of tech-driven inflation all share a common thread – overconfidence in our ability to scale.
The data center industry is no exception. It’s time for policymakers, investors, and industry leaders to take a step back and reassess their priorities. We need to rethink our approach to energy infrastructure, investing in sustainable solutions that can keep pace with demand rather than perpetuating the status quo.
With lead times of three years or more, it’s time for the Big Three to get creative about delivery. They could invest in new manufacturing capacity, prioritize partnerships with smaller players, or even explore alternative technologies like solar or wind power.
Until then, we’re stuck in a cycle of overpromising and underdelivering. The AI revolution will continue to drive innovation, but it’s time for us to get serious about the energy infrastructure that underpins it all – before it’s too late.
Reader Views
- SRSam R. · therapist
The gas turbine shortage highlights the stark trade-off between AI's power hunger and our energy crisis. While the Big Three manufacturers are booked solid, it's worth noting that their production rates can't keep pace with demand. What's more concerning is that these companies are prioritizing high-end clients over smaller players, creating an uneven playing field in a market already plagued by opacity. This has serious implications for industries like renewable energy, which rely on gas turbines to stabilize the grid and ensure a smooth transition to cleaner power sources.
- TSThe Salon Desk · editorial
The gas turbine shortage is more than just a logistical headache - it's a canary in the coal mine for our addiction to AI-driven power. We're so focused on the tech itself that we're neglecting the infrastructure required to fuel its growth. Meanwhile, emerging markets and smaller players are getting squeezed out by the Big Three's prioritization of high-end clients. But what about the real winners here: the investors and companies profiting from this unsustainable energy boom? Their profits may be power-packed, but our future is on shaky ground.
- LDLou D. · communications coach
The gas turbine shortage is just another symptom of AI's insatiable appetite for power. But let's not forget that data centers are also notorious energy hoggers, typically running at 2-5% capacity utilization. The industry needs to wake up to the fact that these behemoths are less about computing and more about cooling – and what a waste of resources it is. We should be investing in edge computing and distributed infrastructure, not perpetuating the status quo with ever-larger data centers.