Is AI Making Current Data Center Infrastructure Obsolete?

In the June 25th Lumen blog, the increased demand AI places on current data center infrastructure is discussed. The Prometheus Hyperscale project in Wyoming (renderings pictured below) is an example of a new data center that will be scaling their power capacity from 120 megawatts to one gigawatt, solving part of the AI power demand challenge. Also, this project’s partnership with SOREDI member Lumen‘s high-capacity, low-latency infrastructure allows them to mitigate the connectivity challenge.

Data centers can be a controversial topic for economic development in some areas due to high power demand, large land requirement and low workforce payroll. For those ready to explore this possibility, however, these centers provide connectivity that stretches across the globe, especially in this everchanging and ever-growing age of AI. Data centers look to be part of the business landscape well into the future until replaced by newer technology yet to be discovered.

AI has become part of our daily lives, including our personal assistant as we navigate the myriads of documents we create in our professional career. Have you ever asked ChatGPT to create a graphic for you? The wait can be up to 10 minutes or more due to the data request load happening at that moment.

Things to consider moving forward: How can we upgrade existing data centers to meet the new high data demand? What are our next steps as we continue to learn how to coexist with AI? How can these centers be scaled down? Other considerations?