AI cloud provider Lambda has teamed up with Cologix to launch Nvidia HGX B200 AI clusters in Columbus, Ohio.

The clusters have been deployed at Cologix's COL4 Scalelogix data center in Columbus, Ohio.

The clusters are comprised of Supermicro's AI-optimized hardware featuring Nvidia Blackwell GPUs, and can be self-served or provisioned through Lambda's GPU flexible commitment model.

“Columbus is a thriving hub for AI innovation, from manufacturing to healthcare,” said Robert Brooks IV, founding team and VP of revenue at Lambda. “With Supermicro’s trusted systems and Cologix’s reliable infrastructure, we’re giving Lambda’s customers in the Midwest the fastest path to production-ready AI—and the added flexibility to integrate with hyperscaler environments.”

Chris Heinrich, chief revenue officer of Cologix, added: “Columbus is one of the fastest-growing digital corridors in the country, and this launch brings coastal-level AI infrastructure into the region.

“Our collaboration with Lambda and Supermicro gives regional enterprises a powerful edge, combining low-latency access, dense interconnection, and ready-to-deploy clusters, giving teams the ability to move faster and scale smarter.”

The Cologix COL4 data center is located at 7500 Alta View in Columbus. The facility spans 256,000 sq ft (23,783 sqm) and offers 33MW of IT capacity across three data halls. Also on the campus are the COL1-3 data centers. In total, the campus spans 500,000 sq ft (46,450 sqm) with 80MW of power available.

Cologix is additionally developing an eight-building campus in Licking County, Ohio. The data center company has said it will invest $7 billion in the project.

Earlier this year, Lambda installed a GB200 NVL72 rack at EdgeCloudLink's hydrogen-powered data center in Mountain View, California.

The AI cloud company raised $480 million in February 2025, with participants in the round including Nvidia, In-Q-Tel, and OpenAI co-founder Andrej Karpathy. It is also a participant in Nvidia's newly launched Lepton AI platform, which connects GPU cloud providers' resources.