The widespread adoption of artificial intelligence is driving increased business investment in data centers, significantly increasing the requirements for backup power, cooling equipment, transformers, and other technologies associated with their operation.
According to the International Energy Agency (IEA), data centers consumed approximately 415 TWh of electricity in 2024, equivalent to 2-3% of global consumption, with an average annual growth rate of 12% over the past five years. The advancement of AI is accelerating the adoption of high-performance servers and increasing the energy density of these facilities. Therefore, the evolution of server and accelerator deployment will be crucial in estimating the future growth of their electricity consumption.
In Mexico, this trend is beginning to translate into concrete infrastructure needs. The Federal Electricity Commission (CFE) and the Mexican Data Center Association (MEXDC) recently agreed on a work plan to address the electrical requirements of 1,705 MW associated with projects at various stages of development. Currently, the industry has 279 MW installed and another 205 MW under construction; by 2031, MEXDC estimates a required capacity of approximately 1.73 GW.
The coordination framework between CFE and MEXDC aims to anticipate the needs of each project, considering its location, power requirements, and projected start date, in order to determine generation, transmission, distribution, and storage requirements in a timely manner. This scenario highlights the importance of incorporating energy planning from the earliest stages of developing new data centers.
Fernando Velasco, general manager of Generac Mexico, noted that AI-related projects represent a significant portion of the company's current orders, ranging from backup generators to solutions incorporating cooling and transformers.
Continue reading
“The increased processing capacity required by these systems is driving up the energy demands of data centers. In Mexico, this trend is already reflected in a greater need for energy infrastructure and equipment with higher capacity, reliability, and efficiency. Therefore, we continue to invest in research and development to meet the needs of this market and support its growth,” he emphasizes.
This trend is not limited to Mexico. In the United States, Generac plans to invest $250 million by the end of 2027 to expand production capacity at several of its plants and meet the growing demand for electrical equipment for AI data centers. This investment is part of the company's ongoing efforts in other markets to address the increased energy requirements associated with these facilities.
The supercomputing and artificial intelligence sectors for data centers have accelerated the purchase of on-site power solutions to an unprecedented level. This growth was one of the topics Generac addressed at DCD Connect Live LATAM 2026, where it participated to analyze the requirements accompanying the expansion of these facilities. The company anticipates that this market will have an increasingly significant impact on its operations and is working on solutions tailored to the needs of each project.
According to Julio Becerril, LatAm Data Center Business Development Manager, to meet this growth, Generac is combining different technologies, from diesel and natural gas generators to battery storage systems and multi-asset solutions. The goal is to provide commercial and industrial customers with greater backup capacity and flexibility to respond to variations in demand. He indicates that, in data centers, where continuity of supply is especially important, this approach is leading the company to expand its offering of generation, storage, control, and energy management solutions.
Solutions developed to meet the needs of data centers include high-capacity power generation platforms, including diesel generators up to 4 MW, as well as integrated solutions that combine battery storage, control systems, and power conversion. These technologies are designed to offer greater scalability, redundancy, and flexibility in hyperscale, colocation, and enterprise installations.