The OpenAI application icon is displayed along with other AI applications on the smartphone.
Jonathan Ra | Nurphoto via Getty Images
Tech giants are turning to nuclear power to power the energy-intensive data centers needed to train and run the large-scale AI models behind today’s generative AI applications.
Microsoft Companies such as Google and Google have agreed to purchase nuclear power from certain U.S. suppliers to provide additional online energy capacity for their data centers.
This week, Google It said it would purchase power from Kairos Power, a developer of small modular reactors, to help “drive the advancement of artificial intelligence.”
“The grid needs these clean, reliable sources of energy to support the development of these technologies,” Michael Terrell, Google’s senior director of energy and climate, said on a call with reporters on Monday.
“We believe nuclear energy can play an important role in helping meet our needs and help meet our needs cleanly and in a round-the-clock manner.”
Google said Kairos Power’s first nuclear reactor will be operational in 2030, and more reactors will be operational in 2035.
The tech giant isn’t the only company looking to nuclear power to realize its artificial intelligence ambitions. last month, Microsoft Signed an agreement with U.S. energy company Constellation to restore an abandoned reactor at Pennsylvania’s Three Mile Island nuclear power plant that has been dormant for five years.
In March 1979, the Three Mile Island Nuclear Power Plant suffered the worst nuclear meltdown and radiation leak in U.S. history, when a valve malfunction caused a loss of coolant, causing the reactor to overheat.
Why they turned to nuclear energy
Technology companies are under pressure to find energy to power data centers – the critical infrastructure behind modern cloud computing and artificial intelligence applications.
Many developers rent servers equipped with GPUs (graphics processing units) from so-called cloud “hyperscalers” such as Amazon, Microsoft and Google, and owning these servers outright is often too costly.
The tech giants have benefited from a surge in interest in generative AI applications such as OpenAI’s ChatGPT. But the increase in demand has also had an unintended effect: energy demand has also increased significantly.
Global electricity consumption in data centers, artificial intelligence and cryptocurrency is expected to double from 460 terawatt hours (TWh) in 2022 to more than 1,000 terawatt hours (TWh) in 2026. International Energy Agency Research Report.
Researchers at the University of California, Riverside, publish A study last April found that ChatGPT consumed 500 milliliters of water for every 10 to 50 prompts it issued, depending on when and where the AI model was deployed. This is approximately the amount of water in a standard 16-ounce bottle.
As of August, more than 200 million people were submitting questions each week on OpenAI’s popular chatbot ChatGPT, according to OpenAI. That’s double the 100 million weekly active users OpenAI reported last November.
environmental opposition
Nuclear energy is not without controversy. Many climate campaigners oppose such supplies on the grounds that they pose dangerous environmental and safety risks and that they do not provide truly renewable energy.
“Nuclear electrodes are extremely expensive, dangerous and slow to build,” climate charity Greenpeace said on its website.
“It is often referred to as ‘clean’ energy because it produces no carbon dioxide or other greenhouse gases when generating electricity, but the reality is that it is not a reasonable alternative to renewable energy.”
Proponents of nuclear energy, on the other hand, say it provides a nearly carbon-free electricity and is more reliable than renewable energy sources such as solar and wind.
“I do think nuclear is the future if it’s built and securitized in the right way,” Rosanne Kincaid-Smith, chief operating officer of global data center provider Northern Data Group, told CNBC at a technology conference in London last week. .
Kincaid-Smith added: “People are scared of nuclear energy because of past disasters. But in the future, I don’t think the traditional grid will be a sustainable energy source as artificial intelligence develops.”
While Northern Data Group is not using nuclear power and has no plans to actively explore using it as a power source for its AI data centers, the company does want to “contribute to this conversation as it has important implications for the broader ecosystem, the broader Ecosystems are very important.” Kincaid-Smith told CNBC.
—CNBC’s Pippa Stevens contributed to this report
About The Author
