AI cloud provider Lambda has signed an agreement with Prime Data Centers to deploy a GPU cluster at a Prime facility in California.

The agreement will see Lambda deploying Nvidia Blackwell AI infrastructure for training and inference workloads at Prime's LAX01 data center in Vernon, California.

LAX01 offers 33MW of capacity across 242,000 sq ft (22,483 sqm), split between six data halls. Under the agreement, Lambda will lease 21MW of capacity.

"Scaling advanced models requires infrastructure built for frontier AI workloads," said Ken Patchett, VP of data center infrastructure at Lambda. "This partnership represents a critical step in Lambda's mission to empower organizations to realize the full potential of AI. This deployment at LAX01 with Prime exemplifies how strategic collaboration can turn ambitious AI visions into operational reality, enabling researchers and enterprises to focus on what matters most: building breakthrough AI applications while Lambda pushes the limits of infrastructure deployment speed, scale, and complexity."

"The race toward superintelligence is reshaping infrastructure requirements across every industry," added Michael Wall, EVP of product delivery for Prime Data Centers. "By combining Prime's purpose-built facilities with Lambda's GPU-dense compute platform, we're working to give businesses the foundation they need to build and deploy the next generation of AI models — faster, more efficiently, and at massive scale."

According to the companies, Prime worked in partnership with Lambda throughout the construction and design process to help speed up the deployment.

Reports that Lambda would be using the site first emerged in August 2024, with Supermicro filing an 8-K with the SEC stating it would lease the capacity and then sublicense the space to Lambda. There was no mention of Supermicro in the latest release.

Prime's current portfolio is largely centered in the US around California, with campuses also in development in Dallas and Austin, Texas; Chicago, Illinois; and Phoenix, Arizona. Prime says it has a 4GW roadmap of power across markets in the US and Europe, and aims to deliver 1GW of capacity by 2028. The company secured investment from Snowhawk and Nuveen in July 2025.

Lambda secured a multi-billion-dollar contract with Microsoft earlier this month through which the AI cloud will give Microsoft access to tens of thousands of GPUs.

Founded in 2012 and offering GPU-based cloud compute, Lambda operates out of 15 data centers across the US, according to its website. The company aims to deploy more than one million Nvidia GPUs and 3GW of liquid-cooled data center capacity.

The company is known to lease colocation space in data centers in San Francisco, California (ECL); Allen, Texas; and Columbus, Ohio (Cologix). EdgeConneX is developing facilities in Chicago, Illinois, and Atlanta, Georgia, for the company. Lambda is set to occupy an Aligned facility in Plano, Texas. It is also set to occupy a facility in Kansas City, Missouri.