Are fuel cells the new favorite for AI power supply? Bloom Energy and Brookfield reach a $5 billion data center power supply project

Asset management giant Brookfield has reached a partnership with fuel cell manufacturer Bloom Energy to provide new energy solutions for AI data centers through the deployment of onsite power generation technology.

According to an announcement released on Monday, Brookfield Asset Management will invest up to $5 billion to deploy Bloom Energy's fuel cell technology. This marks Brookfield's first investment to support the power and computing infrastructure strategy for large AI data centers. Following this news, Bloom Energy's stock price surged over 30% in pre-market trading.

Currently, the massive data center plans in the AI industry are clashing with the aging U.S. power grid, which often struggles to quickly deploy additional power capacity. Bloom Energy's fuel cells provide a rapidly deployable onsite power solution, as they do not rely on the existing grid, effectively bypassing the limitations of traditional power infrastructure.

The two companies stated that they are collaborating globally to design and deliver so-called "AI factories," and will announce a specific project location in Europe by the end of this year. This move signifies that large investors are actively seeking to leverage alternative energy technologies to address the energy constraints faced by AI development.

AI Computing Power Expansion Faces Grid Bottlenecks, Onsite Power Becomes Key

The ambitions of the AI industry are placing unprecedented pressure on the power system. For example, the data center cluster recently announced by NVIDIA and OpenAI has a planned power demand of up to 10 gigawatts, a scale that rivals the entire electricity consumption of New York City during peak summer periods. However, these grand plans are facing a harsh reality: the aging infrastructure of the U.S. power grid makes the process of adding new power capacity slow and complex.

In addition to insufficient grid capacity, the surging power demand from data centers may also exacerbate rising electricity prices for consumers, leading to broader socioeconomic issues. Against this backdrop, finding independent, efficient, and rapidly deployable power sources for data centers has become an urgent priority for tech companies and investors.

As a solution that does not rely on the grid, onsite power generation is becoming key to solving the power supply challenges for AI. Jensen Huang stated last week that the industry needs to build power generation facilities independent of the grid to quickly meet demand and protect consumers from rising electricity prices:

"The self-generation model for data centers may be much faster than connecting to the grid, and we must do this."

Bloom Energy's fuel cell technology precisely meets this demand. By deploying fuel cells onsite at data centers, companies can obtain stable and reliable power supply without waiting for grid upgrades or new transmission lines. This model not only allows for rapid deployment but also provides data center operators with greater energy autonomy and cost control

Brookfield and Bloom's Global Layout

This cooperation is strategically significant for both parties. For Brookfield, this is its first major investment supporting the new strategy for AI data center power and computing infrastructure. The company has been heavily investing in Europe in recent years and earlier announced plans to invest €20 billion to develop AI projects in France, aiming to create the largest AI infrastructure cluster in Europe.

For Bloom Energy, the collaboration with Brookfield is an important milestone for the large-scale commercialization of its technology in the AI field. According to the company, Bloom Energy has previously deployed hundreds of megawatts of fuel cells through partnerships with utility companies such as American Electric Power and data center developers like Equinix and Oracle. This new agreement, worth up to $5 billion, will greatly enhance its business scale and market influence