- The AI boom has triggered a surge in spending on data centers.
- These data centers require billions of dollars of investment and huge amounts of energy.
- It's putting pressure on local utilities and has made rural America a hotbed for building.
It's less than a year since ChatGPT launched to the public, triggering a boom in artificial intelligence investments, and forever changing our comprehension of the technology.
And while the rise of AI has already changed our digital realities, it's also beginning to impact our physical world, too.
The AI boom has supercharged a wave of spending on data centers. This building boom is sucking up billions of dollars, along with water, land, and energy.
"There's a well-publicized arms race happening in AI, and the major tech companies are expected to invest $1 trillion over the next five years in this area, mostly to data centers," Jonathan Gray, Blackstone's president and chief operating officer, said on an earnings call on July 20.
The data center boom is set to double or triple the amount of energy consumed by these data centers.
By 2030, data centers are expected to reach 35 gigawatts of power consumption annually, up from 17 gigawatts last year, according to McKinsey. A recent Cowen research report estimated that AI data centers could require more than five times the power of traditional facilities.
"It's staggering," said Marc Ganzi, CEO of DigitalBridge, an investment firm that owns and operates data centers, fiber networks, and other tech infrastructure. "Not sure how we do it."
Insider has been covering the data center boom at length. Here, we break down what's going on:
AI is supercharging the data center boom
AI model training is especially energy intensive. It requires the use of graphic processing units. These GPUs are specialized chips that multitask better and work faster than central processing units, or CPUs, which run most traditional cloud services.
AI computer servers being installed in data centers are often equipped with multiple GPUs, usually supplied by Nvidia. Each GPU consumes up to about 400 watts of power, so one AI server can consume 2 kilowatts. A regular cloud server uses 300 to 500 watts, according to Shaolei Ren, a researcher at UC Riverside who has studied how modern AI models use resources.
For instance, the power consumption of Nvidia's new GH200 server cluster is about two to four times more than a regular cluster of the same physical size, he estimated.
Tom Keane, who oversaw Microsoft's cloud data centers for about two decades, recently warned about this. Microsoft is the main backer of OpenAI, creator of ChatGPT and builder of the most powerful AI models around.
"In the case of a training data center, you make that as big as possible, you put as many computers in there as possible and you're running that data center at full utilization all of the time," Keane told Bernstein analyst Mark Moerdler in a recent interview. "That physical data center starts to become more resource intensive. You start to design that AI data center very differently."
Why generative AI will have a huge impact on data centers.
Rural America has become a hotbed
On a rolling expanse of rural Ohio land, America's digital future is being sown.
Last year, a partnership between the real estate investors Lincoln Property Company and Harrison Street purchased 190 acres in New Albany, a small city about 20 miles outside of Columbus where the pair plan to begin construction on a 200-megawatt data center by the end of the year.
The project's neighbors include Google, Meta, Microsoft, and Amazon – all of whom have similar plans, or are already underway with major data center projects.
"Our regional message is if you're a major data center developer or customer, we want to talk to you," said Matt McCollister, an executive vice president at One Columbus, a business development group in the region.
The data center industry has long been clustered in a handful of well-established markets, primarily Northern Virginia, Dallas, Phoenix, Silicon Valley, and Chicago. But the emergence of places like New Albany shows how soaring demand and the sector's voracious appetite for energy is increasingly pushing data center developers and users throughout the country.
Artificial intelligence, which requires massive computing power and energy loads, is expected to further this migration – especially as utilities in the industry's core markets have struggled to keep pace with its growth.
Wall Street is betting big
To help it reach $1 trillion in assets, Blackstone has wagered big sums on apartment buildings, warehouses, student housing, and other commercial real estate assets that proved to be shrewd investments.
Now, as borrowing costs rise, property values sink, and a once-soaring real estate market has become perilous, the investment giant is turning to ChatGPT for answers — literally.
Blackstone has been talking up data centers with expectations that the industry will benefit from a boom in artificial intelligence and become a key new area of focus in its $585 billion real estate portfolio.
Two years ago Blackstone funds, including its largest single real estate fund, Blackstone Real Estate Income Trust, known as BREIT, took the data center landlord and developer QTS Realty Trust private in a $10 billion deal that has made QTS the centerpiece of its data center ambitions.
Blackstone executives have said QTS's value has since tripled and that they plan to radically scale it beyond that with $8.5 billion of new data center projects due in the next three years and a $50 billion pipeline of longer-term development.
Inside Blackstone's data center gamble
These complexes are hoovering up huge amounts of energy
Just a few miles from the $5 billion second headquarters that Amazon is raising outside of Washington, DC, the tech giant is in the midst of a far larger, and less conspicuous, building boom.
The company is in the process of developing $87 billion worth of data centers, a push that has already made it the biggest player in the world's largest data center market in northern Virginia. The featureless, warehouse-like structures are easy to miss on the sides of highways or tucked unassumingly amid suburban neighborhoods.
Data centers, including Amazon's, play an increasingly central, but unseen role in modern life, housing the digital infrastructure that powers critical functions such as e-commerce, autonomous vehicles, video streaming, and, now, artificial intelligence.
There is a flipside, however, to their now ubiquitous presence in places like northern Virginia. The facilities consume quantities of power so vast that they have begun to tax entire energy grids and could exacerbate the climate crisis.
Amazon does not disclose how many data centers it occupies, where they are located, or how much electricity they consume. The company's data facilities are tied to its large cloud computing business, Amazon Web Services, which offers software, storage, and other services to legions of customers.
Based on a review of permits that Insider obtained through a Freedom of Information Act request, Amazon operates, or is in the process of building or planning, 102 data centers in northern Virginia. Together, the facilities, when they are all up and running, will have emergency generators capable of producing more than 4.6 gigawatts of power. That's almost enough backup electrical capacity to light up all of New York City on an average day.
That's forcing some utilities to delay their shift away from fossil fuels
In Phoenix and the surrounding region, data centers have attracted attention for the noise they blare, the water they guzzle, and the large tracts of land they've consumed.
Now, the prodigious power burned by this fast-growing industry threatens to overwhelm the city's utilities and stymie efforts to remove fossil fuels from the grid even as the climate crisis has flashed increasingly dire signals. This year, the daily high temperature in Phoenix reached or surpassed 110 degrees Fahrenheit for a record 54 days.
"We have about 7,000 megawatts of data center requests currently in our pipeline," Karla Moran, an executive at Salt River Project, one of two major utilities that serve the Phoenix region, told Insider. Moran noted those requests rivaled the size of the utility's entire 11,000 megawatt system.
While all of those power requests are not expected to materialize into actual development, she said the size of the interest was unprecedented. The data center industry's soaring electrical demand has been so significant, it factored into the utility's far-reaching plans, Moran said.
In October, the power company, better known as SRP, approved a significant expansion of its generating capabilities that includes the development of 2,000 megawatts of new methane gas facilities. Those plants will effectively preserve the size of its portfolio of fossil-fuel-fired infrastructure into the next decade and potentially beyond.
Moran said data centers, with their heavy, around-the-clock power use, were "one of the main reasons we look at having a resource like that."
Data centers are booming. Their need for power is causing utilities to retreat on green energy.