Microsoft’s Fairwater datacenter will use more power than Los Angeles
The next generation of AI data center campuses are city-scale.
Within two years, the largest data centers may use more power than major cities.
We think Microsoft’s Fairwater Wisconsin campus will have a peak electrical load of 3.3 GW when its fourth building comes online in late 2027. Los Angeles, by comparison, used 2.4 GW on average in 2023.
It should be noted that today’s chatbots do not use large amounts of energy per query. Earlier this year, we estimated a typical GPT-4o ChatGPT query uses about as much energy as using a microwave for one second.
It’s unclear how energy-intensive future AI systems will be to run. Increasing use of inference-time compute scaling may push energy usage per query upwards, while algorithmic and hardware efficiency improvements push downwards.
Whatever happens at the model level, one thing is clear: hyperscalers are building enough new infrastructure to put city-scale amounts of power and compute behind AI.
You can find this data insight, along with more analysis about the largest data centers under construction, all on our website.
Big Data is gaining traction among various industries. According to IBM, people produce 2,500 trillion bytes of data daily. 50 billion IoT and other connected devices gather, analyze, and share it (as CISCO states). Big Data unlocks an excellent opportunity for big insights — available for companies of any domain and size.
Big Data should no longer be regarded as an afterthought. Using the information outside your company’s own data sources, it perfectly fits in your business intelligence solutions and expands the comprehension of the market and customers. The companies that will stay afloat — and ultimately lead the pack — will be the ones that put Big Data amongst such priorities as revenue, profitability, and customer experience.
According to the Broadcast Audience Research Council, Big Data solutions contributed to better decision-making (69%), improved customer experience (52%), and significant cost decrease (47%). What’s more, companies bragged about an 8% increase in revenue and a 10% cost reduction because they treated their Big Data properly.
From the way these figures are presented in the post, it sounds like this is comparing the theoretical peak usage of the data center with the average actual usage of the city?
It's an impressive figure no matter how you look at it, but it might be inaccurate to say that the data center will actually use more power then the city?