AI and data center boom drives project cargo

The growth in Data Centers, fueled by AI revolution, is driving a massive demand for electric power. To meet this demand more building of electric generation capacity is needed…and it has already started.

Artificial Intelligence (AI) is driving massive growth in data centers, and consequently an increase in project cargo to serve those facilities.

Over 100 megawatts of new data center construction have been added every month since late 2021 and the US colocation data center market doubled in size in four years, according to AFCOM’s 2025 State of the Data Center Report.

Emerging technologies – especially AI – are the driving forces behind this data center boom. “The vast computational power demands of AI model training are spurring the construction of hyperscale data centers with hundreds of megawatts to more than a gigawatt of power capacity,” says the State of Digital Infrastructure Industry 2025 Report from iMasons.

Consulting company McKinsey says, “Our research shows that global demand for data center capacity could almost triple by 2030, with about 70% of that demand coming from AI workloads.” McKinsey analysis also shows that by 2030, companies will invest almost $7 trillion in capital expenditures on data center infrastructure globally.

McKinsey adds that the growth of AI, and the data centers to host it, has multiple drivers: mass adoption of GenAI, a popular new form of AI that can generate new content based on information collected across massive datasets; the deployment of AI-powered applications across a variety of industries; hyperscalers and enterprises competing to obtain their own proprietary AI capacity; and governments investing in AI infrastructure. The top cloud providers, called hyperscalers; tech industry giants; colocation specialists; major corporations; niche players in AI; crypto companies; and even governments are all participating in the data center boom.

“The race is on right now to build up as much compute capacity as possible,” explains Danny Yunes, VP, Mission Critical Ops and Strategy at Logistics Plus, a 3PL that handles project cargo for data center projects. “Companies are building data centers for AI, and they need power.”

More Power Essential

“AI-ready data centers consume an especially large amount of energy because of their high average power densities — that is, the energy consumption of servers in the racks,” states McKinsey. “Average power densities have more than doubled in just two years and are expected to rise nearly four times by 2027.”

The iMasons report explains further, “AI amplifies the digital infrastructure industry’s power challenge. For example, AI training data centers consume exponentially more power than traditional data center workloads.” And it’s not just the number crunching – the type of graphics processing units (GPUs) needed for AI can cause a server to generate significantly more heat than a traditional server, which means more power is needed for cooling as well. By 2035, Deloitte estimates power demand from AI data centers in the U.S. could grow more than thirtyfold, reaching 123 gigawatts, up from 4 gigawatts in 2024. Similarly, Goldman Sachs estimates that global data center power demand will increase 160% by 2030.

Experts expect that utilities will not be able to keep up with the increasing demand for data center power. Bloom Energy finds that utility providers report significantly longer timelines to deliver power in key US markets, up to 2 years longer than what hyperscalers and colocation providers expect. So, data centers are adopting onsite power systems to bypass or augment the traditional energy grid. According to Bloom Energy, 38% of facilities are expected to use some onsite generation for primary power, and 27% of data centers are expected to be fully powered by onsite generation by 2030, a 27x increase from just 1% last year.

CoreSite, a leading data center colocation provider, explains, “By establishing their own behind-the-meter (BTM) power generation capabilities such as microgrids fueled by gas turbines, fuel cells or modular nuclear reactors, as well as renewable energy sources including solar and wind, data center operators can increase their scalability, flexibility and reliability even as the demand for ever more power grows.”

“The behind-the-meter power concept is very big right now because the traditional power grid is at capacity,” Yunes from Logistics Plus confirms. “So many companies – the hyperscalers and crypto miners for example – have been building their own power generation behind the meter. Billions of dollars are being invested for onsite energy production.”

Moving Parts

The AI-driven data center growth, and the corresponding need for onsite power generation, are energizing the project cargo market. “Anyone who’s building a data center with power generation onsite or near site is going to need to move that equipment to the job site,” Yunes explains.

“We’ve seen an uptick in the number of companies looking for creative ways to create energy off the grid for either compute power or crypto mining,” adds Jason Bernick, VP, Crypto Logistics at Logistics Plus, “and we’re seeing an increase in the amount of project cargo requests to support data centers.”

“Demand is up,” Yunes agrees. “We’re seeing 18-20% growth in the number of project cargo shipments to data center job sites in 2025 compared to last year.”

The type of over-dimensional and heavy equipment Logistics Plus is moving to data centers includes transformers, generators, gas turbines, and cooling towers, with values ranging from $100,000 to tens of millions of dollars. For example, Logistics Plus recently moved a massive modular unit designed for power generation at an isolated off-grid site, weighing roughly 75 tons once loaded on a specialized truck and trailer.

Modular Construction

Another interesting aspect of the data center project cargo wave is the transportation of prefabricated infrastructure modules, manufactured offsite and then assembled onsite to construct the data center building. Logistics Plus is transporting these modules to data center sites as well.

McKinsey says demand for AI-ready data centers is surging, and with it the potential for a serious supply deficit. The supply crunch raises demand for modularized construction, which speeds up the build-out of data centers.

“The modules are like Lego blocks that they build offsite and then ship to a construction site and drop into place. It’s much more efficient construction,” Yunes explains.

While traditional data centers can take 12 to 24 months or longer to complete, modular data centers can be delivered two to three times faster, typically three or four months after the order is placed, according to Wesco, a provider of supply chain services.

“The large building modules are over-dimensional as well, sometimes sent 1000+ miles from origin to destination. We have a number of these new project cargoes lined up,” Yunes verifies.