Data centers must report chemicals that could interfere with treatment, but many consider cooling fluids proprietary. Smaller facilities may not have to disclose them, and municipalities aren’t required to make that information public. Environmental groups also have raised concerns about chemicals used in cooling systems, including PFAS, or forever chemicals. PFAS are linked to health risks, such as certain cancers, as well as reproductive, liver and thyroid problems. While researchers can estimate this indirect water use, Ferreira said, the numbers rely on broad averages and can vary widely depending on the power mix and other factors. Nevertheless, with no special permits required, information on water use is not spelled out publicly, and companies are not held accountable.
How Has Cloud Computing Changed Data Center Energy Consumption?
Modern servers are significantly more efficient than older models, but the sheer volume of computing demand continues driving overall consumption higher. The federal government also has identified data center development as a national priority, committing land and funds to support their growth. GE Vernova (GEV), one of the key suppliers of power equipment and grid infrastructure, just delivered a breakout quarter that underscores how quickly AI is reshaping the energy landscape. Across Wisconsin and the broader Great Lakes region, experts are calling for stronger regulations and a more holistic framework to evaluate data centers before approving them. In several Great Lakes states, competition already exists among municipalities, agriculture, industry and private well owners for groundwater.
Edge computing facilities, designed to bring processing closer to end users, typically consume 50kW to 2MW. While individually smaller, their distributed nature means thousands of these facilities are being deployed globally, contributing significantly to overall consumption. There is no federal registration requirement for data centers, so their estimated number varies depending on the source. Some owners of data centers also obscure their locations for security reasons and/or competitive advantage.
Data Centers Are Not the Villain
Accelerated servers account for almost half of the net increase in global data centre electricity consumption, while conventional servers account for only around 20%. Other IT equipment, and infrastructure (cooling and other infrastructure) demand account for around 10% and 20% of the net increase respectively. All three types of data centres – enterprise, colocation and server provider, and hyperscale – contribute to the growth https://fasthips.com/data-driven-decision-making.html in electricity consumption. Initially, energy concerns in computing were consumer-driven, such as improving battery life in mobile devices. Today, the focus is shifting to environmental sustainability, carbon footprint reduction, and making AI models more energy efficient.
Endangering climate goals and energy infrastructure
Hyperscalers should also clarify their plans to address rising electricity costs, he added. “If they’re quiet about it, that’s going to allow the rumor mill to fly off the handle.” SemiAnalysis argued that PJM’s forecasts often overestimated future demand, particularly as many planned data centers in the area faced construction or assembly delays due to a chronic memory shortage.
Principles for 800 VDC in AI Data Centers: Rack-level Architectures as the Immediate Enabler
This can reduce strain on the grid and water utilities during peak need, making data centers an asset as opposed to a drain. “New electricity consumers such as data centers can actually apply downward pressure on rates by providing utilities more sources of revenue while spreading fixed costs over a larger customer base,” the report says. U.S. data centers consumed more than 4% of the country’s total electricity in 2023, according to the MIT Energy Initiative. The share of these different components in data centre electricity consumption varies greatly by type of data centre, depending on the nature and efficiency of the equipment they have installed. Data centers are now the “number one issue we hear about from our constituents,” says Ian Lovejoy, a Republican state delegate in Virginia.
- Success will require coordinated efforts from industry, government, and consumers to ensure our digital infrastructure supports economic growth while minimizing environmental impact.
- Additionally, the storage and transfer of massive datasets used in AI training require substantial energy, further increasing AI’s environmental burden.
- “As we look ahead to 2026, the rapid evolution of AI continues to redefine expectations for performance, security, and efficiency.
- In 2025, more than 200 bills addressing data center issues were introduced in more than 40 states.
The state’s grid operator predicted peak electricity demand surpassing 367 gigawatts by 2032 — more than four times the current all-time peak of 85,508 megawatts recorded in August 2023. More recently, U.S. electricity consumption has increased since a relative low point in 2020. From 2020 through the end of our short-term forecast in 2026, we expect electricity consumption to grow at an average rate of 1.7% per year. The commercial and industrial sectors grow faster in the forecast, at an average of 2.6% and 2.1% per year, respectively.
Share this:
The governor’s race could be a harbinger of political headwinds for the AI industry’s data center buildout with the mid-term elections just a year away and Democrats zeroing in on affordability as their central issue. In Washington, some Democratic senators are targeting the close relationship that President Donald Trump has developed with the leaders of the major tech companies and AI labs. Meanwhile, voters in localities like Festus, Missouri and Port Washington, Wisconsin have ousted council members or passed referendums to restrict future data center projects. 15 16 These actions highlight the tensions between technological advancement, infrastructure capacity, and local community interests.
- But as the mood toward AI starts to sour, the tides have started turning against data centers as well.
- Electricity prices are surging, voters are growing angry, and the artificial intelligence industry’s data centers are increasingly a target for blame with U.S. mid-term elections on the horizon.
- Plough, of the Protect Fredonia Coalition, said right now it feels like a gold rush, with company leaders vying to see who can get data centers up the fastest, no matter the cost to everyone else.
- In our Annual Energy Outlook 2025 (AEO2025) Reference case, we project the electricity consumed for commercial computing will increase faster than any other end use in buildings.
- Many of the projections and commitments made by AI companies to use clean electricity for their data centers have “gone out the window,” he said.
- In the United States, data centers consumed 4.4% of total electricity in 2023 and could reach 6.7-12% by 2028.
- Mortenson’s Maja Rosenquist and CBRE’s Gordon Dolven here discuss how AI’s growth has accelerated data center development, how site-selection strategies are evolving, and the challenges posed by power constraints.
- Although a few network operators have achieved 100% renewables (including BT, TIM and T-Mobile), data transmission network operators generally lag behind data centre operators in renewable energy purchase and use.
- However, in the wider context, a 3% share in 2030 means that data centre share in global electricity demand remains limited.
- However, they also create challenges including increased electricity costs for other consumers and strain on local infrastructure.
- Companies choose data center locations based on a variety of factors, including the availability of capable power utilities, properly zoned land, and high-quality network access.
As a result, the same level of demand for digital services and AI is met with a reduced electricity consumption footprint. This unlocks energy savings of more than 15%, with global electricity demand from data centres reaching around 970 TWh by 2035. This energy demand has grown at roughly 12% annually over the past few years, significantly outpacing demand growth in other sectors. As artificial intelligence and cloud-based services scale, projections indicate that data center electricity demand could reach nearly 3% of total global consumption by 2030.
The report noted how NVIDIA continues to lead the AI server market, which has a 95% market share. Last year, NVIDIA shipped 100,000 units that consume an average of 7.3 TWh of electricity annually. States are looking at a variety of methods to ensure that data centers bear the full cost of their https://canadatc.com/technology-for-applying-venetian-plaster-stages.html grid impact.
United States: The Global Leader
Data centers are experiencing unprecedented global growth, due to the explosive development of artificial intelligence and cloud computing. They are used to house IT infrastructure – the servers, storage systems and hardware on which our digital culture stands. In the US, natural gas provides more than 40% of the electricity for data centers, according to the International Energy Agency.
Electric Vehicle Charging Infrastructure: Stations, Equipment, and Planning
Looking ahead, some projections indicate that AI could drive data centers to consume 2,500 to 4,500 TWh of global electricity by 2050, equivalent to 5-9% of total global electricity demand. The Department also provides technical assistance to support states, utilities, grid operators and technology developers right-size the grid in the midst of demand growth through the Supercharging the Electric Grid effort. The combination of rapidly growing size of models and computing demand are likely to outpace strong energy efficiency improvements, resulting in a net growth in total AI-related energy use in the coming years. Although AI itself can help reduce energy use in data centres, the rapid and mainstream adoption of AI chatbots like OpenAI’s ChatGPT and Google Bard are likely to accelerate growth in energy demand for AI.