Artificial intelligence deals with energy demand at the data center – it is only worse

Using artificial intelligence energy Actually, up to 20 percent of global energy demand, Thursday publishing research In the Jul offers magazine. This request from Amnesty InternationalThe United States, it can double by the end of this year, which includes nearly half of the total data center electricity Consumption around the world, with the exception of the electricity used Bitcoin Mining.

The new research was published in the comment of Alex de Fis Gao, the founder of Digiconomist, a research company that assesses the environmental influence of technology. De VRIES-GAO, a classification expert in late 2010, started to explore the effect of bitcoin mining, which is another very thick activity for energy, on the environment. He says that looking at artificial intelligence has become more urgent over the past few years due to its wide dependence of Chatgpt and other large language models that use huge amounts of energy. According to his research, energy demand has now been set around the world to overcome the demand from bitcoin mining by the end of this year.

“The money that bitcoin mines had to reach where they are today are Sudanese beans compared to the money that Google and Microsoft “All these large technology companies flow (to artificial intelligence),” this is escalating much faster, and it is a much greater threat. “

The development of artificial intelligence already has an impact on Big Tech’s climate targets. Their technology giants Acknowledging in recent sustainability reports Artificial intelligence is largely responsible for increased energy use. For example, greenhouse gas emissions from Google increased 48 percent since 2019, complicating the company’s goals of reaching zero by 2030.

“Since we integrate artificial intelligence into our products, reducing emissions may be difficult due to the increased energy requirements of more intensity to calculate artificial intelligence,” Read.

Last month, the International Energy Agency issued a report Finding databases constitute 1.5 percent of global energy use in 2024-415 hours, slightly less than the annual Saudi energy demand. This number has only been set to become larger: Electricity consumption has grown four times faster than total consumption in recent years, while the amount of investment in data centers has almost doubled since 2022, largely driven by the tremendous expansion of the new artificial intelligence capacity. In general, IEA expected the electricity consumption to grow in the data center to more than 900 TwH by the end of the contract.

But there are still a lot of unknown people about the participation that artificial intelligence takes specifically, the current composition of the use of electricity by data centers. Data centers operate a variety of services-such as hosting cloud services and providing online infrastructure-which are not necessarily related to intense energy activities in artificial intelligence. Meanwhile, technology companies maintain the energy expenses of their programs and special devices.

Some attempts to measure the consumption of artificial intelligence energy began by the user: Calculating the amount of electricity that is included in one chatGPT, for example. De VRIES-GAO decided, instead, in the supply chain, starting with production to obtain a more global image.