New AI Technology Uses Whopping Amounts Of Power

Todd Moses
July 6, 2024

Google announced this week it is well behind on a pledge to eliminate net carbon emissions by 2030.

Inputs that matter: Google said the jump highlighted "the challenge of reducing emissions" while it invests in the build-out of large language models and their associated applications and infrastructure.

  • Google admitted that "the future environmental impact of AI" was "complex and difficult to predict."
  • Chief sustainability officer Kate Brandt said the company remained committed to the 2030 target but stressed the "extremely ambitious" nature of the goal.
  • "We still expect our emissions to continue to rise before dropping towards our goal," said Brandt.
  • CNBC reported that electricity demand is forecast to grow by 20% by 2030, with AI data centers alone expected to add about 323 terawatt hours to the US's electricity demand.

The opportunity: As tech giants including Google, Amazon, and Microsoft have outlined plans to invest tens of billions of dollars in AI, climate experts have raised concerns about the environmental impacts of power-intensive tools and systems.

  • Meanwhile, energy generation and transmission constraints are challenging companies seeking to develop new technology.
  • Analysts at Bernstein said in June that AI would "double the rate of US electricity demand growth, and total consumption could outstrip current supply in the next two years."
  • These large-scale data centers use GPUs that are enormously heat-producing.
  • The water used to cool these GPU chips is freshwater, coming from the same reserves used for drinking water.

Zoom in: Google's data center electricity consumption increased 17% in 2023 and amounted to approximately 7-10% of global data center electricity consumption, the company estimated.

  • Google said that its data centers also consumed 17% more water in 2023 than in the previous year.
  • Microsoft reported in May that its total carbon emissions have increased by nearly 30% since 2020, primarily due to the construction of data centers.
  • ChatGPT alone uses half a million kilowatt-hours per day.
  • Amazon has been "the world's largest corporate purchaser of renewable energy for four straight years."

Between the lines: While renewables will likely play an essential role in meeting AI energy demands, analysts say immediate implementation is challenging.

  • The AI push concerns more brownouts in Texas, California wildfires, stronger Gulf hurricanes, and 126 degrees in Delhi this spring.
  • Data centers, which house computer systems and associated components, had already been burgeoning with the Internet and cloud computing.
  • As the value of cryptocurrency has multiplied, so have the data centers, such as those that mine Bitcoin in cheap energy havens like Plattsburgh, New York.

Follow the money: AI's voracious electricity consumption is driving an expansion of fossil fuel use, including delaying the retirement of some coal-fired plants.

  • However, AI is not just a problem but also a potential solution.
  • It is already being harnessed to make the power grid smarter and speed up the innovation of new nuclear technologies.

Go deeper: Subscribe to the free newsletter to learn more.

Read More