Venture capitalists have significantly increased their investments in AI startups, committing over half a trillion dollars to the sector within the past five years.
However, a report from Sightline Climate suggests that the most strategic investment related to AI today might lie within the energy sector. Their research indicates that up to 50% of newly announced data center projects could face delays, with power access identified as a primary obstacle.
Out of the 190 gigawatts of data center capacity tracked by Sightline, merely 5 gigawatts are currently under construction. While approximately 6 gigawatts of projects within Sightline’s database became operational last year, a substantial 36% of projects originally slated for 2025 have experienced timeline setbacks. These delays are anticipated to eventually impact large enterprises and other businesses relying on AI for their operations.
This growing imbalance between supply and demand presents a significant opportunity for investors.
Major technology firms, including Google and Meta, are dedicating substantial portions of their financial resources to developing solar, wind, and nuclear energy projects. Furthermore, these companies are actively supporting innovative technologies, such as Form Energy's 100-hour battery, through direct investments and collaborations with utilities to expedite their integration.
A multitude of startups are emerging to address this power challenge. For example, Amperesand, DG Matrix, and Heron Power are innovating new power conversion technologies, while firms such as Camus, GridBeyond, and Texture are developing sophisticated software solutions to manage the flow of electricity efficiently.
Power availability continues to be one of the most critical limitations for data centers, a situation not expected to improve in the near future. Goldman Sachs projects that AI will escalate data center power consumption by a staggering 175% by 2030.
The current grid shortages are unparalleled in contemporary history, contributing to escalating electricity prices nationwide. This trend has compelled numerous technology companies to investigate alternative methods for powering their data centers. (Notably, the Trump administration, anticipating a potential political challenge, has encouraged tech companies to develop their own power sources, accept higher rates, or both, though many were already pursuing such strategies.)
Leading tech giants including Amazon, Google, and Oracle are actively working to reduce their reliance on the conventional power grid. Consequently, several new data centers are being designed to incorporate either fully on-site power generation or a hybrid model combining on-site power with a grid connection.
The largest data centers are at the forefront of this transition. While less than a quarter of projects that have secured a power source intend to utilize on-site or hybrid solutions, these projects collectively account for a significant 44% of the total planned capacity.
This paradigm shift is partly attributed to shortages of key power generation equipment, particularly gas turbines, alongside an aging and outdated grid infrastructure. This situation has paved the way for the adoption of alternative energy sources.
Google’s recent agreement to power a new data center in Minnesota exemplifies one innovative strategy to address this challenge. The company plans to integrate wind and solar energy with a colossal 30 gigawatt-hour battery system from Form Energy. Additionally, Google collaborated with Xcel Energy to develop a novel rate structure designed to foster the integration of new technologies within the utility’s planning framework.
Form Energy’s battery solution is not an isolated case; grid-scale battery systems are poised to capture a substantial share of the power market. According to the U.S. Energy Information Administration, the U.S. is projected to achieve nearly 65 gigawatts of battery storage capacity by the end of this year. Capitalizing on this momentum, Form Energy, like many of its counterparts, is actively raising a $500 million funding round in anticipation of a future initial public offering.
However, energy supply represents only one facet of the challenge. Once electricity reaches the grid or a data center, it requires sophisticated management, a role traditionally fulfilled by the humble transformer.
The majority of modern transformers still rely on a 140-year-old technology involving large iron cores wrapped in copper wire. While dependable, this design is proving increasingly cumbersome as data center power demands surge. An expert informed TechCrunch that when server racks achieve 1 megawatt of power density, the necessary power equipment will consume double the physical space of the rack itself.
This has prompted a recent surge in investor interest in solid-state transformer startups, which aim to replace the antiquated iron-and-copper technology with silicon-based power electronics. Although currently more expensive than traditional transformers, their inherent flexibility allows them to consolidate the functions of several pieces of data center equipment, thus enhancing their cost-competitiveness.
Overall, the magnitude of investments flowing into battery and transformer companies has remained considerably smaller compared to the blockbuster funding rounds witnessed in the AI industry.
This disparity is not necessarily a drawback; these smaller rounds often present more manageable investment opportunities. Furthermore, as global electrification expands across sectors from transportation to heavy industry, the demand for power is set to continuously increase, offering investors a robust hedge against a potential downturn in the AI market. It suggests that perhaps the most astute investment related to AI isn't directly within artificial intelligence itself.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.