ATLANTA — Microsoft CEO Satya Nadella announced Wednesday that the company was unveiling a new, ‘purpose-built’ data center in Atlanta. The data center first started operations in October.
Known as Fairwater 2, the Atlanta facility will be paired and connected with the first Fairwater site in Wisconsin.
Nadella said in his announcement that the move would be part of a “broader Azure footprint to create the world’s first AI superfactory” and was an example of building infrastructure that “can serve any workload, anywhere,” no matter what the need is.
[DOWNLOAD: Free WSB-TV News app for alerts as news breaks]
The tech giant said to meet the “unprecedented demand for AI computing,” it had reinvented how it approached data center design.
The Fairwater facilities are “a departure from the traditional cloud data center model and uses a single flat network,” integrating hundreds of graphics cards to process data as a massive supercomputer.
TRENDING STORIES:
- Data center set to use 6 water towers worth of water per day in Georgia
- Anthropic announces $50B investment in new US data centers to meet AI demand
- DeKalb County to hold town hall on proposed data center regulations
The new facility designs used by Microsoft are also intended for training future models and multiple different workloads with diverse requirements.
“A traditional datacenter is designed to run millions of separate applications for multiple customers,” Alistair Speirs, Microsoft general manager focusing on Azure infrastructure, said in a statement. “The reason we call this an AI superfactory is it’s running one complex job across millions of pieces of hardware. And it’s not just a single site training an AI model, it’s a network of sites supporting that one job.”
Microsoft said the plants feature liquid-based cooling, individualized power to each server rack, where the data is stored and computed, as well as facility-wide cooling systems to ensure long-term viability, while also being more energy efficient.
Another feature of the Atlanta Fairwater facility’s design is its approach to resource management. Microsoft officials said the advanced liquid cooling system “consumes almost zero water in its operations” while maximizing the amount of processing chips it can host and use.
“We are pushing the envelope in serving this compute with cost-efficient, reliable power. The Atlanta site was selected with resilient utility power in mind and is capable of achieving 4×9 availability at 3×9 cost,” the company said in a statement. “By securing highly available grid power, we can also forgo traditional resiliency approaches for the GPU fleet (such as on-site generation, UPS systems and dual-corded distribution), driving cost savings for customers and faster time-to-market for Microsoft.”
[SIGN UP: WSB-TV Daily Headlines Newsletter]
©2025 Cox Media Group
