AI will require even more energy than we thought

6 Min Read

It’s no secret at this point that popular generative AI tools like OpenAI’s ChatGPT have a high need for data. The billions and sometimes trillions of information parameters needed to train these models are housed in massive data centers that use electricity for cooling and processing power. But new predictions and forecasts suggest that increasing demand for increasingly powerful AI models could stretch the current energy supply further than we ever thought. In the US alone, according to a new report released by the Electric Power Research Institute (EPRI) data centers tasked with powering advanced AI models could be responsible for 9.1% of the country’s total energy demand by the end of the decade. Much of that new demand could be met by non-renewable natural gas, which could complicate global efforts to reduce carbon emissions.

The EPRI analysis warns that widespread adoption of generative AI tools in the coming years could result in a “step change in power requirements.” By 2030, the report said, data center energy needs could account for between 4.6% and 9.1% of total U.S. electricity production in 2030. That’s compared to 4% now. The renewed demand is not limited to the US either. By 2026, the International Energy Agency (IEA) will estimates Demand for energy in data centers could double globally by 2026.

Much of that predicted increase in demand, the report notes, will come from unique energy-intensive generative AI models. EPRI estimates that a simple search for OpenAI’s ChatGPT uses about 10 times as much electricity as a typical Google search. That large disparity is likely due to the enormous amount of training data and computing power required to make these models perform as intended. And that’s just for text responses. The amount of data generated by emerging generative AI audio and video models such as OpenAI’s Sora “has no precedent,” according to the report. One thing seems clear: AI is responsible for powering the growing energy needs of data centers. A recent one prediction According to a publication by financial giant Goldman Sachs, AI alone will be responsible for 19% of data center power needs by 2028.

See also  Samsung Galaxy Watch 7 could get much faster charging times

Data centers are turning to fossil fuels to meet short-term energy needs

Energy-hungry data centers could put great pressure on energy networks in the coming years. According to Goldman Sachs’ forecast, data centers will be responsible for between 1 and 2% of global energy demand from 2024. This figure is expected to rise to 3-4% by the end of the decade. In the US, which operates approximately half of the world’s data centersthese facilities are expected to be responsible for 8% of the country’s total energy needs by 2030. are energy suppliers already rushing to bring new power plants online to ensure brewing energy demand is met. Goldman Sachs forecasts estimate that more than half (60%) of the energy used to meet this demand will come from non-renewable resources. This forecast reinforces previous reports suggesting that renewable sources alone may be insufficient to meet the energy needs of data centers.

[ Related: Sam Altman: Age of AI will require an ‘energy breakthrough’ ]

Also the new energy needs further complicate previous explanations from technology leaders like OpenAI’s Sam Altman, who have suggested that powerful AI models could play a long-term role in reducing greenhouse gas emissions. Altman, who previously said the era of powerful AI would require an “energy breakthrough,” was Reportedly among a handful of prominent Silicon Valley figures who recently invested $20 million in Exowatt, a startup that seeks to harness solar energy to power AI data centers.

But data centers and energy providers don’t necessarily have to wait for a technological silver bullet to address some of AI’s energy dilemmas. In its report, EPRI called on data centers to explore ways to increase internal efficiency by reducing the amount of electricity used for temperature cooling and lighting. Cooling only, Reportedly responsible for approximately 40% of data center energy consumption. Additionally, EPRI notes that backup generators powered by renewable energy sources can also play a role in supporting more reliable, sustainable energy networks.

See also  Information operation targeting Ukrainian speakers in the context of the war

“Shifting the relationship between the data center and the power grid from the current ‘passive load’ model to a collaborative ‘shared energy economy,'” the report notes, “could not only help electric utilities cope with the explosive growth of AI, but also contribute to affordability and reliability. for all electricity users.”

Share This Article
Leave a comment

Leave a Reply

Your email address will not be published. Required fields are marked *