Patmos unveils 10MW AI data center in converted Kansas City site

Patmos has announced the expansion of its custom-built artificial intelligence (AI) data center facility in downtown Kansas City, alongside the immediate availability of 10MW of new colocation space for high-density GPU, high-performance computing (HPC), and AI infrastructure.

The new facility spans 360,000 square feet and has been purpose-built to accommodate the demands of AI and advanced computing workloads. It represents Patmos' first dedicated AI campus and marks a transformation of the former Kansas City Star newspaper printing press building, known locally as "The Star", into a modern data centre environment.

Facility details

This development increases the total local capacity at the Patmos data center to 35MW, enabling AI companies and enterprises to rent rack-ready colocation in increments of 2.5MW. According to Patmos, the facility can be built out to client specifications, with new white space made operational within a 90-day turnaround period. Power densities exceed 100kW per rack, supporting the requirements for high-density, next-generation AI infrastructure.

The site incorporates in-building cooling plants, including three cooling towers, and uses district distilled water systems, which the company claims contribute to sustainability and green energy objectives. The model of retrofitting historical structures in urban areas reportedly allows Patmos to deliver new data centre capacity more quickly than building new facilities, providing data center space in a matter of months rather than years.

Operational focus

The project includes a focus on both technology and community. Workspaces for start-ups and established enterprises are incorporated alongside areas for conferences, co-working, and social events. This is intended to foster a hub for technology professionals and organisations in the city's downtown district and to support the ongoing revitalisation of the area.

"The demand for AI infrastructure far outstrips the available supply, and Kansas City - like many major cities in the Midwest - is ideally placed to meet this demand in the U.S., with the skilled workers, the connectivity and the appetite from municipal authorities to make it happen," said Joe Morgan, Patmos' Chief Operating Officer.

Morgan outlined the need for rapid deployment of new AI infrastructure as a market priority with Patmos taking a different approach through retrofitting existing industrial spaces.

"Building AI-ready data centers from scratch is a complex and time-consuming process, and going to hyperscalers for new infrastructure can take years. By retro-fitting data centers out of facilities like the Star building, the Patmos AI Campus approach can deliver the capacity needed in months, not years, and with 10MW of new capacity available now we're excited to welcome new tenants to the facility.

"Meanwhile, we're creating the social and co-working spaces that will make this facility valuable for humans, not just servers - which is not just a 'nice to have', it's a requirement that helps revitalize the area, create jobs, and make high-tech work for people not just corporations."

Security and sustainability

Security at the facility includes 24x7 armed staff and an artificial intelligence-enabled robot patrolling the premises. Patmos asserts that such measures, combined with the choice to use district resources and the focus on energy efficiency, further enhance the suitability of the facility for large-scale data workloads and enterprise operations.

Company representatives stated that the flexible design and redevelopment approach of properties such as "The Star" can support a growing tech workforce in downtown Kansas City. The ongoing transformation is expected to generate new jobs and stimulate local economic activity while supporting the needs of technology firms and AI research initiatives.

The full 10MW of new data center capacity is currently available to AI and HPC organisations requiring large-scale and rapid deployment options, with leasing available in 2.5MW increments. The facility continues to expand its operational scope and tenant services, as it aims to accommodate a growing share of nationwide demand for AI and high-density computing infrastructure.