Artificial Intelligence, Apps, and Websites Find New Home in New York

ORANGEBURG, New York — Apps and websites are seen as things that exist solely in cyberspace, but the online world is sustained by a lot of physical infrastructure. This infrastructure is increasingly in the spotlight amid the rise of power-hungry AI technologies like ChatGPT.

Some of that physical infrastructure takes the form of large buildings full of computer equipment, known as data centers. Now, about 25 miles north of Manhattan, one of these key data centers is being built on a construction site in Orangeburg, New York.

“This is a purpose-built data center that will have five suites, five data halls,” says Dan Fuentes, senior vice president of corporate sales at DataBank, the company behind the project.

“Watt” is a data center?

Databank’s Orangeburg location is the size of a warehouse. But in the industry, data centers aren’t measured in square footage, but in megawatts of power.

“It will have 20 megawatts of critical power for our customers, or 30 megawatts total,” Fuentes tells ABC Audio.

Databank operates much like an office building owner. Companies as large as Apple and as small as neighborhood banks contract with the company, and those companies configure and maintain their own computers and servers. Databank, meanwhile, is responsible for ensuring that those devices are always safe, cooled, and powered.

“So we have large customers, large hyperscalers, large banks, large enterprise customers, AI customers who want to put their infrastructure in a place where they know it will never go down. So these data centers are purposely built to never go down,” Fuentes says.

Uninterrupted power supply

The data center is being built next to an electrical substation, from which the center will draw power most of the time, according to Fuentes. But in the event of a power outage or natural disaster, massive generators that are set up outside the building kick in.

Angel Otero, the data center’s operations manager, says that in the event of a major power outage, the generators can still provide power as long as they have fuel, something he encountered at another data center during Superstorm Sandy in 2012.

“Every refinery on the East Coast was underwater, crippled,” he says. “We had a steady supply of fuel from Philadelphia. And we had – we just ran on generator power for about a week. No problems, no worries, no trouble. As long as we had fuel, we were up.”

A growing industry

According to the U.S. International Trade Commission, there are currently more than 8,000 data centers operating around the world, with more than 2,800 located in the United States alone.

“So globally, there are about 40 gigawatts of data center infrastructure installed worldwide. About half of that is in the United States,” says Databank CEO Raul Martynek.

And he says business is booming.

“All of these computations, as this technology becomes more advanced, they use more compute cycles, they use more bandwidth, they use more storage. And all of that is driving demand for data centers,” Martynek says.

McKinsey predicts the data center industry will grow by 10% by the end of the decade. And while companies like Meta, Amazon, and Google have their own data centers, they also rent space in purpose-built facilities like the one DataBank manages.

Cliff Stein, a professor of industrial operations studies and computer science at Columbia University, says all this interest in data centers shouldn’t come as a surprise.

“Almost everything we do in the world is somehow connected to—either we use them directly or we use something that is needed in these data centers. So to keep the economy going, we need that,” Stein says.

Everything from social media companies to streaming services rely on data centers, and recently generative AI has entered this market.

How AI is Impacting the Data Center Industry

“With the big language models, everything took off and evolved at an incredible pace,” Stein says.

Martynek says the rapid rise of AI-powered tools like ChatGPT, Google Gemini, and Apple Intelligence has spurred even greater growth in his industry.

“It’s kind of like before and after ChatGPT. There’s been a tsunami of demand in the last two years because everyone wants to adopt this technology and incorporate it into their business. I think the industry has been really caught off guard,” Martynek says.

That’s why, he says, the sector is now in a race to build new data centers.

Environmental Impact of Data Centers

At the Orangeburg facility, Fuentes says, their customers’ computer equipment will be housed in one of five 20,000-square-foot rooms known as “data halls.” The floors will be installed about 3 feet off the ground, leaving room for industrial-strength air conditioners. Because all that computing can generate a lot of heat, the design allows for cool air beneath the rows of computers.

Otero believes this means it can be noisy inside the data center.

“Depending on how much the customer is using at any given time, yes, it can get pretty loud in there. Noise-cancelling headphones are definitely a must,” he says.

In addition to noise, air conditioners also require a lot of power. When you combine this with the normal energy demands that such a data center requires, some environmental questions arise.

“Using a lot of energy, depending on the source, affects the climate in the same way that more cars would,” says Jake Bittle, a journalist at Grist, who covers climate change and energy.

“If we’re using a lot of gas and coal-fired power plants to run these data centers, that’s going to have an impact on the climate because more fossil fuels are being burned to power them,” he says.

The AI ​​boom has only added fuel to the fire. A Goldman Sachs Research report found that asking ChatGPT a single question uses as much energy as ten Google searches. According to MIT Technology Review, training a large language model generates as much CO2 as five gasoline-powered cars. As more data centers spring up to meet the demand, all that energy use could show up on consumers’ electricity bills, Bittle says.

“You’re probably going to see some kind of strain on the overall power supply. It may not have happened yet, but that will lead to some kind of impact on the ratepayers, who are, you know, residential and commercial businesses,” Bittle says.

Martynek says that to counteract this, the industry is investing in more environmentally friendly energy sources.

“We know our power grid is a bit outdated, it hasn’t been modernized. And we need more renewable energy connected to the grid,” he says.

However, both Stein and Bittle are skeptical that the data center industry will rely entirely on renewable energy sources, at least in the near future.

“I think it might make a difference, but it won’t solve the problem completely,” Stein says.

“I think that while companies are doing a lot of good things to reduce energy and water usage in the data centers themselves, they are just energy-hungry, and that has consequences,” Bittle says.

Fuentes says AI companies only account for a fraction of the total energy sold by Databank.

“AI is a component of that, but it’s tiny compared to the spectrum of verticals we sell to.”

But the technology will be present at Orangeberg. Databank says that when the location has its grand opening later this year, 80% of the facility will be leased to Coreweave, an AI infrastructure company.