Microsoft Unveils Fairwater AI Datacenter in Wisconsin
Microsoft has unveiled Fairwater, a 315-acre AI datacenter located in Mount Pleasant, Wisconsin. This facility is part of a broader strategy to enhance AI infrastructure and is designed to deliver ten times the performance of today's fastest supercomputers. The datacenter spans 1.2 million square feet and houses hundreds of thousands of NVIDIA GPUs, making it one of the most advanced AI facilities globally.
Fairwater is equipped with a closed-loop liquid cooling system to manage its high computing density, ensuring zero water waste. The facility's architecture allows it to process up to 865,000 tokens per second, significantly boosting AI model training capabilities. NVIDIA GPUs are interconnected in large clusters, enabling parallel AI training and high-bandwidth communication.
The datacenter represents a significant investment in AI infrastructure, with Microsoft planning to spend over $7 billion in Wisconsin alone. This includes a second facility of similar scale, expected to be operational within three years. The project has created thousands of construction jobs and will support hundreds of full-time positions once operational.
Microsoft's investment in AI datacenters like Fairwater underscores its commitment to leading the high-performance computing race, with similar facilities planned in Norway and the UK. These efforts are part of a global strategy to build a robust AI infrastructure network, positioning Microsoft at the forefront of AI development.
We hope you enjoyed this article
Consider subscribing to one of our newsletters like Silicon Brief or Daily AI Brief.
Also, consider following us on social media:
More from Data Centers
Oct 16 EverWind Plans 420 Megawatt AI Data Centre in Nova Scotia Oct 16 Gartner Says $3.5 Trillion in AI Infrastructure Spending Faces Impairment Risk by 2030 Oct 16 KBRA Rated Anthropic BB+ Before $35 Billion Chip Financing Oct 16 China Proposes Export Controls for Advanced AI Models and Chips Oct 16 Hitachi Energy and Ansaldo Energia Partner on Data Center Power SystemsSilicon Brief
Weekly coverage of AI hardware developments including chips, GPUs, cloud platforms, and data center technology.
Whitepaper
Tensordyne Napier: What If One Rack Could Do the Work of Nine?
Tensordyne
This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.
Read moreYou may also like
USD.AI Arranges $128.9 Million GPU Financing Facility
Brightray Signs Letter of Intent for 10 MW Indiana AI Data Center
NVIDIA Commits $250 Million in Compute to UK AI Safety Research
Everpure Adds AI Data Management and Inference Tools
Forge Commits $30 Million to Deploy Private AI Systems With webAI
Daily AI Brief: the AI news that matters, in your inbox.