Private Equity Investment in the Picks-and-Shovels of the AI Infrastructure Boom

Written By Kevin Zhou

Data centers and the digital infrastructure that supports them have become the modern-day picks-and-shovels of the artificial intelligence (AI) revolution. Private equity (PE) is capitalizing on the opportunities this creates with massive investments. In 2024, PE-backed data center M&A reached C$18.15 billion globally—the highest total in at least five years, according to S&P Global Market Intelligence.

S&P also reports that private equity is prioritizing investments in the infrastructure that underpins AI's rapid expansion. In May 2025, Blackstone stated that with data centers, they are investing in the picks-and-shovels of the digital economy, and are well positioned as the long-term demand for compute goes up.

Here is a look at five rapidly evolving areas of digital infrastructure that form the backbone of the AI boom.

Cooling Systems

Running data centers with less energy is one of the most pressing challenges in the AI era. According to the US Department of Energy, as much as 40 percent of data centers' energy consumption goes towards cooling computing hardware.

Heating, Ventilation and Air Conditioning (HVAC) systems play an indispensable role in maintaining optimal operating temperatures in data centers. Key components of HVAC systems include computer room air conditioning units, air handler units, chilling and cooling towers, air distribution systems and precision sensors.

Liquid cooling, while more complex than traditional air-based systems, absorbs significantly more heat and results in lower energy consumption. It also supports higher-density hardware configurations. It involves using liquid as the primary medium to absorb and dissipate heat generated by high-performance computing equipment, such as servers, Graphics Processing Units (GPUs) and other hardware.

Forbes notes that while liquid cooling may still be categorized as “emerging,” it is quickly becoming essential infrastructure for AI data centers. In June 2025, Amazon Web Services announced its transition from an air-based to a liquid-based cooling solution.

Automation and Monitoring

Automation in AI data centers leverages software and systems to perform tasks with minimal human intervention. This plays a critical role in energy optimization, predictive maintenance and workload management. McKinsey identifies AI-driven automation as a key area of investment for hyperscalers, colocation providers and GPU-as-a-service platforms.

Internet of Things (IoT) devices are equipped with sensors and other technologies and have the ability to streamline operations and reduce costs. According to IoT Business News, advanced applications of IoT in data center infrastructure include real-time monitoring of temperature, humidity and air quality, continuous data collection, cost optimization, energy efficiency and enhanced cybersecurity.

Fiber Optic Networks

High-speed fiber optic networks are essential for transferring large datasets quickly and efficiently, both within and between data centers. PitchBook says that fiber optic networks are one of the critical tools of next-generation computing.

The CEO of Corning said in a recent interview that data centers need 10 times more fiber for first-generation AI, and with each new generation of GPUs, this grows by four times. Fiber is also becoming even more innovative to fit higher volume connections in the limited physical space of data centers.

Data Storage and Processing

High-density storage systems are designed to store vast volumes of data in a compact physical footprint. They ensure fast data access and processing, while optimizing space and energy efficiency in data centers, which is crucial as data demands grow.

Edge computing deploys computing resources closer to the end-users or devices generating data. It is particularly useful in AI applications that require low-latency responses, such as security systems, IoT devices and real-time analytics. Edge computing also allows for localized processing even if the connection to the central data center is disrupted.

GPUs and Tensor Processing Units (TPUs) are powerful chips designed to accelerate processing and are a cornerstone of AI data centers.

Secure Facilities

Data centers must be designed to protect the facilities from unauthorized access, theft and physical damage. Common features include surveillance systems, biometric access controls and perimeter protection. Secure facilities are essential for ensuring compliance with regulatory standards.

To discuss any of these pick-and-shovel opportunities in the digital infrastructure that supports AI, please contact Kevin Zhou.

Bennett Jones Private Equity & Investment Funds Practice

The Bennett Jones Private Equity and Investment Funds group is a leader in Canada. Our clients include sophisticated financial sponsors who are looking to balance risk with expected return and who require tailored advice from the initiation of the investment phase to exit. Bennett Jones represents all sides in private equity transactions, with particular depth acting for both US and Canadian sponsors.

Download PDF