8 US Data Center Operators Driving AI Infrastructure

AI is accelerating data center development and career progression across the US, pushing operators towards larger campuses, denser compute and more complex power, cooling and network requirements.

The eight companies below show how that buildout is taking shape, from gigawatt-scale campuses and high-density GPU infrastructure to interconnection, power strategy and modular AI systems.

1. QTS Data Centers

QTS is expanding across major US data center markets including Virginia, Georgia, Texas, Arizona, Ohio and Oregon.

A significant part of that growth is the infrastructure connecting its campuses. QTS is working with Lumen to connect 16 new US locations to high-capacity fibre infrastructure designed to support AI, cloud and hyperscale traffic.

That matters because large AI environments do not operate as isolated buildings. They rely on fast links between compute, storage, cloud platforms and other facilities, making network capacity part of the wider infrastructure challenge.

Projects of this type can involve expertise across:

• Data Center Development
• Electrical and Mechanical Engineering
• Construction Management
• Commissioning
• Critical Facilities
• Fibre Infrastructure
• Network Engineering
• Capacity Planning
• Utilities and Power

QTS provides a good example of how large-scale data center construction and network infrastructure are becoming increasingly connected.

2. Microsoft

Microsoft’s planned campus in Pecos, Texas, demonstrates the scale some US AI infrastructure projects are now reaching.

Announced in June 2026, the project is expected to add approximately 2GW of data center capacity over a five-to-seven-year development programme. Microsoft expects more than 6,000 construction jobs at peak, followed by hundreds of permanent operational roles.

A development of this size involves far more than building data halls. Power generation, substations, electrical distribution, civil infrastructure, cooling and network connectivity all have to be delivered alongside the compute environment.

Infrastructure at this scale can involve roles across:

• High-Voltage Electrical Engineering
• Substations
• Civil Engineering
• Mechanical Engineering
• HVAC and Cooling
• Construction Management
• Electrical Trades
• Pipefitting
• Controls
• Commissioning
• Data Center Operations

The Pecos project is a useful example of how much of the AI infrastructure buildout depends on engineering, construction and skilled trades as well as technology.

3. Vantage Data Centers

Vantage is developing its Frontier campus in Shackelford County, Texas, specifically around large-scale AI infrastructure.

The 1,200-acre development is planned to include ten data centers, more than 3.7 million square feet of space and 1.4GW of critical IT capacity.

The technical design is particularly interesting. Frontier is being developed to support racks exceeding 250kW and will use liquid cooling for next-generation GPU workloads. That level of density changes what is required from the facility around the hardware. More power has to be delivered into a smaller space, while significantly more heat has to be removed without compromising reliability.

Infrastructure designed for this type of compute can involve:

• Liquid Cooling
• Mechanical Engineering
• Chilled Water Systems
• Thermal Engineering
• High-Density Electrical Design
• Controls and BMS
• Commissioning
• Critical Facilities
• Construction Management
• Power Infrastructure

Vantage shows what AI-ready infrastructure increasingly means in practical terms: not simply space for GPUs, but power and cooling systems designed around much higher densities.

Explore Jobs

4. CyrusOne

CyrusOne’s current expansion in Texas shows how closely data center development is becoming tied to access to power.

Its planned development near the Freestone Energy Center includes an initial 380MW agreement, with another 380MW planned for a second phase.

Across its Texas developments, CyrusOne has more than 1.1GW of power under contract. Locating major capacity alongside energy infrastructure can reduce one of the biggest constraints facing new data center projects: getting sufficient power to the site quickly enough. Projects built around this type of strategy can involve expertise across:

• Utility Coordination
• High-Voltage Engineering
• Transmission
• Substations
• Electrical Design
• Power Systems
• Energy Infrastructure
• Construction
• Commissioning
• Critical Facilities

As AI demand increases, power strategy is becoming part of the data center development process much earlier than it once was.

5. Google

Google is expanding its US data center footprint alongside major investment in the energy infrastructure needed to support it.

In Texas, Google and Intersect are developing the Meitner Energy Center, where new data center capacity is being developed alongside new energy generation in Gray and Roberts Counties. The approach brings power and compute development closer together rather than treating electricity supply as a separate problem to solve once the data center is planned.

Google is also developing new data center capacity in other US markets, with projects increasingly connected to new generation, energy storage and grid investment.

In Minnesota, for example, its Pine Island development is linked to plans that include 1,400MW of wind, 200MW of solar and 300MW of iron-air battery storage on Xcel Energy’s system. Alongside those physical projects, Google is also changing the way its data centers use electricity.

By March 2026, the company had integrated 1GW of data center demand-response capacity into long-term agreements with US utilities. That allows some machine-learning workloads to reduce or shift demand when the grid needs greater flexibility.

Infrastructure of this type can bring together:

• Data Center Development
• Electrical Engineering
• Grid and Utility Infrastructure
• Energy Storage
• Power Systems
• Controls
• Data Center Operations
• Capacity Planning
• Infrastructure Software

Google’s current projects show how closely new data center capacity is becoming linked to power generation, storage and grid planning.

6. Equinix

Equinix operates more than 70 data centers across the US, with facilities in major markets including Northern Virginia, Silicon Valley, Dallas, Chicago and New York.

Its infrastructure model is different from the large greenfield AI campuses being developed elsewhere. Equinix is heavily focused on colocation and interconnection, allowing customers to connect directly to cloud platforms, carriers, networks and other businesses from within its facilities.

That makes connectivity as important as the physical space itself.

Infrastructure across environments like these can involve:

• Critical Facilities Engineering
• Data Center Operations
• Network Engineering
• Fibre
• Cloud Interconnection
• Data Center IT
• Electrical and Mechanical Operations
• Controls
• Customer Engineering

Equinix highlights another important part of the AI infrastructure market: the interconnected facilities that allow compute, cloud services and networks to work together.

7. AWS

AWS is taking another approach to deploying AI infrastructure through its AI Factories model. The model allows dedicated AWS AI infrastructure to be installed within a customer’s own data center. The customer provides the facility and power, while AWS supplies and operates the compute, networking and storage environment.

That infrastructure can include AWS Trainium accelerators and NVIDIA GPUs alongside high-performance network and storage systems.

It creates a model that sits between conventional private infrastructure and the public cloud.

Deployments of this type can involve:

• GPU Infrastructure
• Data Center IT
• High-Performance Networking
• Storage
• Systems Engineering
• Infrastructure Deployment
• Power and Cooling Integration
• Cloud Engineering
• Security
• AI Infrastructure Operations

AWS shows how the physical data center and cloud infrastructure are becoming more closely connected as AI environments become more specialised.

8. ECOBLOX

ECOBLOX takes a different approach again, focusing on modular AI and high-performance computing infrastructure rather than relying solely on traditional data center development.

Its modular environments combine compute, power, cooling and infrastructure management into systems designed to bring AI capacity online more quickly. ECOBLOX says deployments can be operational in approximately three to six months, depending on the project.

The infrastructure can support both air and liquid cooling and is designed around high-density AI hardware, including NVIDIA and AMD accelerators. Its approach ranges from retrofitting AI infrastructure into existing buildings to deploying dedicated modular data center units.

Projects of this type can involve:

• GPU and HPC Infrastructure
• Liquid Cooling
• High-Density Power
• Modular Data Center Engineering
• Network Infrastructure
• DCIM
• Deployment Engineering
• Commissioning
• Data Center Operations

For organisations that need compute close to available power, existing facilities or specific workloads, modular infrastructure provides a different route to adding AI capacity. Hamilton Barnes recently entered a strategic partnership with ECOBLOX, combining our specialist recruitment expertise with its work across modular AI infrastructure.

What Is Changing Across the US Data Center Market?

The projects above show that US data center growth is not following one model.

Microsoft and Vantage are developing AI infrastructure at enormous scale. CyrusOne is bringing data center development closer to energy infrastructure, while Google is linking new capacity more closely to generation, storage and grid flexibility.

QTS is investing in the fibre connecting major campuses. Equinix remains focused on the interconnection layer between networks, cloud platforms and enterprise infrastructure. AWS is bringing dedicated AI systems into customer facilities, while ECOBLOX is using modular infrastructure to shorten deployment times.

Across those models, several areas are becoming increasingly important:

• Power and High-Voltage Infrastructure
• Data Center Construction
• Liquid Cooling and Thermal Engineering
• Commissioning
• Critical Facilities
• Fibre and Network Infrastructure
• GPU and HPC Infrastructure
• Data Center IT

For professionals working in these areas, the growth of AI infrastructure is creating more routes into the data center market than the traditional operations path alone.

Building a US Data Center Team?

Hamilton Barnes recruits across Data Center Construction, Critical Facilities, Commissioning, Data Center IT and AI Infrastructure throughout the US.

If you’re building a team, speak to our Data Centers & AI specialists about the skills you need and where that talent sits.

Explore Data Centers & AI