Nuclear Energy Makes a Comeback as Tech Companies Fuel the Next Generation of Power Plants

The biggest tech companies in the world, including Google, Amazon, and Microsoft, are turning to nuclear energy to meet the growing demands of their artificial intelligence (AI) operations. As the energy requirements of their AI systems continue to skyrocket, they are finding that traditional sources of electricity are not sufficient to power their data centers, which are both energy-intensive and require a constant supply of power. The Conversation interviewed numerous experts in the field, and the overwhelming consensus is that nuclear energy is now the best option for these tech giants.

Key Takeaways:

  • The world's largest tech companies are increasingly relying on nuclear energy to power their AI operations due to the high energy demands of their data centers.
  • The energy intensity of AI systems is staggering, with a single Google search using as much electricity as turning on a household light for 17 seconds, and a single question asked of a Generative AI model like ChatGPT using as much electricity as leaving a light on for 20 minutes.
  • Nuclear energy is a reliable source of power that can meet the constant needs of data centers, unlike renewable energy sources like solar and wind which provide intermittent power.
  • The cost of building a 1.1 GW nuclear power facility is estimated to be around $7.77 billion, although this can vary depending on the location and other factors.
  • Tech companies are incentivizing power providers to invest in nuclear energy by guaranteeing they will purchase electricity far into the future, effectively mitigating the risks associated with long-term forecasting.
  • Nuclear energy is also becoming more affordable, with the cost of building a single nuclear reactor now more palatable compared to the expense of AI data centers, which are both more costly and entirely useless without electricity.
  • The first phase of the OpenAI and SoftBank's Stargate AI project will cost $100 billion and could be entirely powered by a single nuclear plant, highlighting the potential for a new generation of tech-enabled nuclear power plants.

Statistics:

  • The estimated cost of building a 1.1 GW nuclear power facility is around $7.77 billion.
  • The construction time for a 1.1 GW nuclear reactor is around 7.5 years in the US and 6.3 years globally.
  • Nuclear energy produces only six tonnes of CO2 per gigawatt-hour of electricity, compared to coal which produces 970.
  • Nuclear energy is also among the safest forms of energy, with per gigawatt-hour causing 820 times fewer deaths than coal, 43 times fewer than hydropower, and roughly the same as wind and solar.
  • The total costs of the Fukushima disaster in Japan are estimated to be around $188 billion.

Sources:

  • The Conversation -- Canada -- By Goran Calic, Associate Professor of Strategy and Entrepreneurship Leadership Chair, McMaster University
  • https://theconversation.com/ai-is-consuming-more-power-than-the-grid-can-handle-nuclear-might-be-the-answer-258677
  • World Bank
  • SoftBank
  • OpenAI