Technology companies like Microsoft, Amazon and Google want to support the rapid growth of generative AI. looking for renewable energy sources.

It is estimated that by 2026, data centers will consume more than 1 terawatt (a trillion times a watt) of electricity per hour. This is more than what existing data centers need. 160% more. Therefore, it is necessary to close the gap somehow.

We cannot keep up with the speed of artificial intelligence in producing energy.

Very large language models, deep learning and cloud-based data processing; serious energy It requires. For example, large AI models such as OpenAI’s GPT-4 are trained on billions of parameters.

Training these systems can consume megawatts of energy. So even if you only train one large AI model, annual energy consumption of a medium-sized city equivalent It can happen.

According to a study, a single ChatGPT query can change a light bulb Equal to the amount of electricity needed to burn for 20 minutes consumes large amounts of electricity.

The energy sources we have cannot provide the energy needed for artificial intelligence.

Solar and wind energy are not sufficient for the continuous and large-scale energy needs of artificial intelligence. Every power outage, billions of dollars in damage can cause. Therefore, there is a need for a source that provides uninterrupted energy.

This is where nuclear energy comes into play. Power stations; regardless of the weather conditions, with low CO2 emissions and high efficiency can work.

As you can imagine, tech giants; artificial intelligence To meet the ever-increasing energy demand with nuclear energy it works.

Google’s agreement with Kairos Power by 2030 It is thought that Amazon and X-Energy will be ready by 2039.

If you would like to view our news on this topic: