Explosive electricity demand in the AI era: Addressing it with SMRs
Just as the Industrial Revolution of the 18th century was triggered by the invention of the steam engine powered by coal, any new technological or industrial development requires an energy source. The AI ecosystem is expanding, and operating generative AI will require an enormous amount of electricity. For example, a simple Google search consumes only 0.3Wh of electricity, but conducting the same search using generative AI consumes about 10 times more power.
At the World Economic Forum (WEF) on August 17th, Toshiaki Higashihara, Executive Chairman of the Hitachi Group, stated during his keynote speech that as AI becomes more activated and widespread, power demand at the data centers will surge, potentially increasing electricity consumption by up to 1,000 times by 2050.
While the argument that AI could lead to an energy crisis has been around for years, the emergence of Open AI's ChatGPT in November 2022 and the rapid expansion of the AI revolution has drastically changed the situation. To operate advanced AI, computers must solve complex calculations quickly, consuming a massive amount of electricity and generating heat, which also requires substantial energy to cool. In response, big tech companies are building data centers. These centers, due to their extensive electricity and water usage for cooling, are placing a focus on “heat management solutions” as a key technology in the AI era for cost efficiency. Solving the rapidly increasing power demand while achieving carbon neutrality is a challenge that will significantly impact both big tech companies and national competitiveness.
Nuclear fusion is the energy source that major countries and large ICT corporations are focusing on to solve this problem. Nuclear energy and nuclear fusion, which is mentioned by French President Macron and CEO Sam Altman at the Davos Forum, is seen as the most practical solution.
Given the rising awareness of rapidly changing climate conditions, it is challenging to find alternatives other than nuclear and fusion power to quickly meet the increasing electricity demand. In 2021, Altman personally invested $375 million in Helion Energy, a U.S. nuclear fusion company. Helion Energy signed a contract with Microsoft in May last year to supply fusion energy.
In the AI era, Small Modular Reactors (SMRs) are emerging as a proposed solution to meet the explosively growing electricity demand. On August 8th, Robert Eccles, a professor at Harvard Business School, highlighted three key advantages of SMRs: “base load characteristics,” “small plant site area,” and “plant lifespan.”
When comparing the operational efficiency rates of different energy sources, nuclear energy stands at 92%, which is significantly higher than natural gas, coal, hydro, and wind power. Moreover, nuclear energy requires much less land area compared to renewable energy sources. It occupies 31 times less space than solar power and 173 times less than wind power.
In line with this, China has completed the world's first SMR, Linglong 1, and the European Union launched the European SMR Industry Alliance earlier this year with the goal of operating SMRs by 2030. Both regions are heavily focused on SMR completion and utilization, competing to lead this effort. Korea is also planning to complete the development of its own i-SMR by 2028. The government is set to approve standard design by 2028 and will begin commercializing it by 2035.
It is time to intensify efforts to meet the surging electricity demand while securing safe energy sources. <저작권자 ⓒ 먼데이타임스 무단전재 및 재배포 금지>
|
Most Popular
최신기사
|