The AI Race Is Creating a New Growth Market for Manufacturers

Artificial intelligence may be powered by software, but the AI race is increasingly a race to build physical infrastructure.

Every new AI workload requires more compute capacity. That means more servers and semiconductors, but it also means more power generation and distribution, thermal management, networking equipment and other complex infrastructure to keep those systems operating reliably.

The scale of the buildout is extraordinary. McKinsey estimates that meeting global demand for data center capacity could require $6.7 trillion in investment through 2030. Our analysis of that research shows that much of this investment is concentrated in engineered systems and technology, including building automation, power and thermal infrastructure and compute systems, where complex-product manufacturers play a critical role.

At the same time, the demands placed on that infrastructure are changing. AI workloads are contributing to higher rack densities, growing cooling requirements and new power architectures. Data centers are becoming more complex, multi-domain systems built on fragmented global supply chains, increasing the importance of modular design, simulation and supplier collaboration.

For manufacturers, this creates an opportunity that is easy to overlook:

You don’t have to be a data center company to benefit from the data center boom.

The data center boom is becoming a manufacturing boom

Headlines about AI infrastructure tend to focus on the hyperscalers building massive data center campuses. They can’t do it alone, especially at the pace they’re moving.

Behind every facility is an ecosystem of companies designing and manufacturing thermal management equipment, power systems, building automation, servers, racks, chips, networking infrastructure, controls and countless other components.

There is particularly strong alignment between the market and manufacturers of thermal management systems, power equipment, servers and IT systems, and semiconductors.

That's important because many of the capabilities these markets suddenly need are not new.

Industrial companies already know how to engineer pumps, heat exchangers, power systems, controls and complex mechanical equipment. Electronics and high-tech companies already know how to develop increasingly dense, software-enabled hardware.

What has changed is the market around those competencies.

AI infrastructure is creating new demand for them at far greater speed and scale.

How can you use your capabilities to solve an important data center problem?

That leads to a different strategic question for manufacturers.

Instead of asking, How do we become a data center company? leaders should be asking:

Which capabilities that already differentiate us solve an important data center problem?

Consider thermal management. High-density AI and HPC environments are increasing the need for airflow management, liquid cooling, containment, rack-level thermal systems and hybrid cooling architectures. Manufacturers developing those systems must balance thermal performance, energy efficiency, footprint, reliability and serviceability while accommodating customer-specific configurations.

A manufacturer with deep expertise in fluid systems, heat exchangers, pumps, enclosures or industrial cooling may therefore possess capabilities that suddenly have greater value in the data center market.

Power tells a similar story. AI power requirements and grid constraints are increasing interest in modular power blocks, backup power, energy storage and other standardized platforms that can accelerate deployment.

And within electronics and high tech, existing expertise in servers, semiconductors, PCB design, packaging, networking and configurable hardware maps directly to increasingly complex AI infrastructure.

The competitive opportunity isn't necessarily about developing an entirely new competency.

It's about applying an existing competency to a rapidly growing problem before your competitors do.

AI readiness starts with product data: 5 trends shaping the future of E&HT

the companies best positioned for the future are building connected product-data foundations

See the Trends

Entering the market is the first step, scaling is the challenge

The speed of the data center buildout raises the bar for how quickly manufacturers can design, validate, configure, manufacture and support those products.

Thermal management makes that complexity particularly visible. Manufacturers may need to coordinate mechanical, electrical, controls and fluid or thermal design across assemblies containing fans, pumps, piping, manifolds, sensors, heat exchangers and control systems. A change to one component can affect performance models, bills of material, validation plans and service information.

Meanwhile, customer requirements vary. Products may need different configurations for different facility layouts, operating conditions, regional requirements or cooling architectures.

This is where a strong product development systems become a competitive advantage.

A manufacturer doesn't simply need to innovate once. It needs to turn that innovation into a repeatable, configurable, and serviceable product platform.

The AI infrastructure race requires a digital engineering backbone

PTC's core technologies address different pieces of that challenge across the product lifecycle.

CAD: Turn engineering expertise into products faster

For manufacturers developing complex data center equipment, CAD is more than a documentation tool. Design and simulation provide the environment in which teams can explore new architectures and validate performance earlier.

PTC's Creo, for example, supports the design, simulation and optimization of complex cooling assemblies, modular platforms, enclosures, airflow paths and liquid-cooling architectures.

For companies adapting an existing product portfolio to data center requirements, that capability can help engineering teams explore new configurations without starting from scratch.

PLM: Make complexity scalable

As product families expand, manufacturers also need to control the growing web of product data surrounding them.

Windchill provides the PLM foundation for managing product data, configurations, supplier inputs and engineering changes across teams. Within data center applications, PLM can connect design, sourcing and operations while providing greater control over configuration and change.

That matters when growth depends not on delivering one specialized product, but on creating a platform that can be configured repeatedly for different customers.

ALM: Connect requirements, software, and hardware

Data center products are also becoming increasingly software-enabled.

Codebeamer supports requirements, risk, testing and validation processes for complex engineered systems, helping manufacturers maintain traceability across engineering and lifecycle processes.

That connection becomes increasingly important as hardware, controls and software evolve together.

SLM: Extend the opportunity beyond delivery

Finally, the business opportunity doesn't end when equipment arrives at the data center.

Cooling, power and other infrastructure become long-lived assets that require service, maintenance and replacement parts. ServiceMax supports service planning, warranty support, spare-parts management and installed-base visibility for these complex systems.

Taken together, the value proposition is straightforward:

CAD helps create the product. PLM governs its complexity. ALM connects its requirements and software. SLM helps keep it performing in the field.

Your next growth market may already be within reach

The AI infrastructure boom will create winners far beyond the companies building AI models or operating hyperscale data centers.

It is creating an adjacent growth market for manufacturers with exactly the engineering expertise the physical AI economy requires.

The manufacturers positioned to capitalize on this moment will be those that identify where their capabilities solve an important infrastructure challenge and build the digital foundation required to iterate, configure, manufacture and service those solutions at data center speed.

AI is reshaping the data center—and it could reshape the growth map for your business.

The question isn't necessarily whether your company can become a data center company.

It's whether the capabilities you've spent decades building can give you an advantage in one of the most significant infrastructure buildouts of the coming years.