"At the end of computing power lies electricity"—this saying circulating in the tech circle is becoming a reality faster than imagined. AI data centers are springing up like mushrooms, and their insatiable appetite for electricity is driving demand in the gas turbine industry.

Recently, Jereh Group, a major domestic player, announced that its wholly-owned subsidiary has secured an order for gas turbine generator units worth USD 301 million (approximately RMB2.08 billion), specifically for powering a data center for a U.S. client. This marks the company's sixth contract of the same kind signed in the United States within six months.

As the global AI arms race intensifies, a parallel competition for electricity is underway.

Hungry AI Devours Electricity

The evolution of AI is also a sprint in energy consumption.

Training a large model like GPT-4 consumes enough electricity to power tens of thousands of households for a year. And that's just the beginning. When the model is put into practical use, processing billions of queries daily, its continuous power demand truly tests the limits of the power grid.

According to the International Energy Agency, global data center electricity consumption accounted for 1%-1.3% of total global electricity use in 2022, and this proportion is accelerating amid the AI boom. Tech giants are not only racing to release models with larger parameters but also "hoarding" land globally to build larger data centers.

The question is, where will the electricity come from?

The construction cycle of traditional power grids is measured in years, while the expansion of AI data centers is measured in months. An even more challenging issue is that in many ideal locations for computing hubs, the grid capacity is already saturated. This has made electricity the biggest bottleneck for cutting-edge tech development.

As a result, the market is turning its attention to gas turbines, which can be rapidly deployed and provide independent power supply. It may seem hard to believe that in today's era dominated by new energy, gas turbines—a relic of the traditional energy era—have become the "hot commodity" for AI data centers.

The answer lies in the unique needs of AI data centers: speed, flexibility, and reliability.

First, speed is essential. From planning to operation, time is money and the key to seizing market opportunities. Gas turbine generator units feature modular designs that can be quickly transported and assembled on-site, with deployment cycles measured in months, perfectly matching the "can't wait" pace of AI projects.

Second, flexibility is crucial. AI business growth is uncertain, and computing demand can explode suddenly. Gas turbines offer adjustable unit sizes, allowing for phased capacity expansion based on demand, avoiding the financial pressure of heavy upfront investments.

Finally, reliability is non-negotiable. Data centers require power stability at the "five nines" (99.999%) level. Gas turbines can form independent microgrids, independent of external grid stability, acting as a "fuse" to ensure uninterrupted core computing power.

Jereh Group revealed in its announcement that its products, with standardized and modular designs, can "effectively address real-world challenges such as tight project timelines and limited space," winning customer favor. This is not just about selling equipment but providing a "plug-and-play" energy solution.

China's Strength Goes Global

The surge in demand has instantly ignited the global gas turbine market.

It is understood that Siemens Energy, an industry giant, has gas turbine orders scheduled for delivery until 2030. The contradiction between booming demand and limited capacity highlights the industry's heat and tension, opening a window of opportunity for companies with technical reserves and manufacturing capabilities.

Jereh Group's RMB 2 billion order announced yesterday is just a microcosm of China's integration into the global high-end industrial chain. This contract is the sixth order Jereh has secured in the U.S. market since November 2025, with clients expanding to five. Its product performance, delivery capabilities, and service systems are gaining sustained recognition in North America, the world's largest data center market.

According to the contract, the equipment will be delivered by the end of 2027. The payment terms are highly favorable to the supplier: the client pays a deposit upfront, with the remainder paid in progress installments and settled in full before shipment. This is not only a testament to technical strength but also an embodiment of industrial chain influence.

Jereh Group's RMB 2 billion order is just a signal of how AI is reshaping energy demand. AI data centers, as a new form of massive, high-stability energy consumption, are changing the investment logic and construction models of energy infrastructure. Distributed, rapidly deployable gas turbine solutions may become one of the "standard configurations" for future large-scale computing clusters.

For China's manufacturing industry, this track represents not only an opportunity for export earnings but also a chance to secure a foothold in global energy technology transformation by leveraging its engineer dividend and supply chain advantages. From the reality of "electricity chasing computing power," we see a new path for China's high-end equipment to take center stage on the global arena.