The Zhitong Finance App learned that Calvin Butler, CEO of Exelon Corp (EXC.US), said that the development of artificial intelligence is expected to drive up to 900% increase in electricity demand in data centers in the Chicago region. According to Butler, about 25 data center projects are currently underway at its subsidiary Commonwealth Edison Co. Engineering research is carried out within the scope of business. The total electricity consumption of these projects is expected to reach 5 gigawatts, which is equivalent to the power generation capacity of 5 nuclear power plants. Currently, the data center's demand for this system is approximately 400 megawatts, and 80% of the planned development projects are nearing completion.

Butler added: “We have seen extensive activity from data center developers who are attracted by ComEd's low electricity prices, carbon-free power generation from nuclear reactors, and national tax incentives.” An Exelon spokesperson stressed that the company will continue to ensure that its systems can handle additional power load growth.

Currently, power companies and regulators are facing the biggest increase in demand in decades. In addition to supporting data centers that run artificial intelligence calculations, the US grid is also preparing for new plants and electrification transitions, such as the spread of vehicles and heat pumps.

Almost all organizations — whether private companies, public institutions, universities, or IT service providers — rely on data centers to store and benefit from their digital data and IT systems, such as cloud computing services, hosting websites, databases, and applications.

However, data centers require significant amounts of electricity to power their computer servers and cooling systems. According to the International Energy Agency, data center electricity consumption accounts for an average of 2% of the country's electricity consumption. This level of consumption is expected to rise further as data processing and storage requirements continue to grow.

Especially today, when artificial intelligence technology is booming, the need for electricity is even more urgent. Billionaire Elon Musk notes that although the development of artificial intelligence was “limited by chips” last year, the key bottleneck has now turned into “electricity supply.”

Amazon CEO Andy Jassi also warned earlier this year that “the current energy supply is insufficient to support a new generation of generative artificial intelligence services.” According to reports, Alphabet, the parent company of Amazon, Microsoft, and Google, is investing billions of dollars to enhance its computing infrastructure to meet the huge demand for resources from artificial intelligence, especially data centers that usually require years of planning and construction.

However, some popular data center construction sites, such as Northern Virginia, are facing capacity limitations, which are further driving the global data center market's growth in the search for suitable locations.

“Demand for data centers has always been high, but we may not have enough capacity to meet global demand in 2030,” said Kaji Sharma, executive vice president of Schneider Electric's data center division. Sharma's team is currently working with chipmaker Nvidia to design a center optimized for AI workloads.

Daniel Golding, a former Google data center executive and chief technology officer at Appleby Strategy Group, mentioned, “In the new artificial intelligence economy, one limitation of deploying chips is where we build data centers and how we get electricity. To some extent, the electricity supply will limit the further development of artificial intelligence.”