Home / Business and Politics / Microsoft invests over $10 billion in green energy due to ‘hungry AI’

Microsoft invests over $10 billion in green energy due to ‘hungry AI’

In 2020, Microsoft promised to become carbon neutral by 2030 by offsetting carbon and reducing its own emissions. Additionally, the tech giant committed not only to being net positive for the climate but also promised to remove from the environment the carbon that the company emits directly or indirectly through electricity consumption. But then came energy-hungry artificial intelligence, and it requires a lot of energy.

Artificial intelligence is already consuming monstrous amounts of electricity, changing the equation for tech companies balancing between dual priorities. These are managing emissions and staying competitive in the AI race. Data centers that store information and host large language models on which artificial intelligence is trained are equipped with high-voltage computing chips that require a lot of energy and a lot of water for cooling.

Some analysts predict that artificial intelligence could consume as much as a quarter of all energy in the U.S. by 2030. For instance, Microsoft revealed last September that its water consumption had increased by 34 percent. The existing electricity grid is not capable of handling such a large increase in demand, so tech companies are taking matters into their own hands.

And Microsoft is doing just that, having recently announced a $10 billion contract with asset manager Brookfield to supply Microsoft with a massive 10.5 gigawatts of renewable energy capacity between 2026 and 2030. This is by far the largest such contract ever signed, and the amount of gigawatts is equivalent to the amount of electricity needed to power four million homes.

– Energy companies used to be excited about contracts for four megawatts. Then they turned into deals of 600 megawatts. And now we are in gigawatts. We see all of this growing at an exponential rate – said Xavier Smith, director of research, energy, and industry at the research firm AlphaSense.

Brookfield is a giant in the renewable energy space. It manages over 7,000 power generation facilities across five continents, from wind farms and solar plants to hydroelectric plants. But even Brookfield will likely need to build even more capacity to meet Microsoft’s demands.

This could come in the form of more solar projects, which could fit well given the proximity of potential solar sites in the Sunbelt to the growing data center hubs in Phoenix and Las Vegas.

However, it is not just renewable sources that are on the radar of tech companies; nuclear energy is also emerging as a lifeline, as Amazon last month purchased a nuclear-powered data center in Pennsylvania, while OpenAI CEO Sam Altman backed a startup aiming to power data centers with small nuclear reactors.

– Nuclear energy is increasingly becoming an important part of this conversation. Just last fall, Microsoft had a job advertisement for a nuclear technology manager. They want someone who can develop a strategy for implementing small modular reactors – says Smith.

No company has specified what types of renewable energy Brookfield will supply to Microsoft or the precise financial terms of the contract. The Financial Times estimates that the deal is worth around 10.5 billion dollars, based on current industry trends and pricing structures. Renewable energy prices have significantly dropped in recent years and could continue to do so, but Smith says it still makes sense for companies like Microsoft to enter into contracts now rather than wait for the possibility of cheaper energy in the future or try to build their own energy infrastructure. As artificial intelligence rapidly evolves, unlike energy projects that can take years, tech companies also need to adapt quickly.

Microsoft operates over 300 data centers worldwide, which it uses to store information and train large language models that power chatbots like ChatGPT and AI applications including Copilot. Due to the distributed nature of AI infrastructure, some have suggested that off-grid, customized energy projects that directly power data centers could be a more efficient way to supply electricity for AI. While it might be cheaper for Microsoft or its competitors to build their own off-grid solutions, Smith said that this is a less attractive option.

– Although tech companies want to be exposed to renewable energy sources, they do not want to manage a power plant. Just as Google does not want to run a power plant, Microsoft does not want to operate a nuclear facility – concludes Smith.

Tagged: