Newly established neocloud provider Kidz AI has signed a Memorandum of Understanding (MoU) with Limestone Networks to lease colocation space at the company’s Dallas, Texas, data center, creating the physical foundation for its next GPU deployment. The agreement centers on Limestone’s DFW3 facility, where Kidz AI plans to deploy an Nvidia B300 GPU cluster designed to serve enterprise inference customers. The move reflects a strategy that ties infrastructure investments directly to contracted revenue rather than future expectations.

The proposed deployment covers 1.125MW of capacity, although the company continues evaluating two potential layouts within the facility. One option places the entire deployment inside a dedicated data hall, while the second distributes infrastructure across two halls if connectivity, proximity, and power requirements remain aligned. If Kidz AI proceeds with a dedicated hall, the rollout will begin with approximately 0.5MW to 0.6MW before expanding to the full 1.125MW as GPU servers arrive and installation progresses. This phased approach allows computing capacity to grow alongside hardware availability instead of requiring a full-scale launch on day one.

Long-Term Customer Contracts Shape Infrastructure Investment

The Dallas deployment directly supports Kidz AI’s previously announced 60-month GPU services agreement valued at $44.6 million with inference platform Canopy Wave. Rather than acquiring infrastructure ahead of customer demand, the company has linked physical expansion to an existing commercial commitment. As a result, the agreement illustrates a business model focused on utilization certainty before capital deployment, an increasingly important consideration as GPU infrastructure costs continue rising across the AI sector.

“Securing data center capacity is the next step in executing our $44.6 million agreement with Canopy Wave,” said Stephanie Luo, CEO of Kidz AI. “We are not pursuing capacity on a speculative basis. We secured long-term contracted demand first, and we are now putting in place the infrastructure required to serve it. That sequencing – demand first, capacity second – is the core of our neocloud model. “Equally central is our strategic focus on inference compute: we are purpose-building this cluster for high-throughput, cost-efficient inference on open-weight AI models, which we believe is where enterprise demand is growing fastest. Together, we believe this approach supports disciplined capital allocation, greater visibility into utilization, and a repeatable playbook for subsequent GPU clusters and enterprise compute agreements.”

Catalyst Compute Leads GPU Cluster Deployment

Kidz AI has assigned its subsidiary, Catalyst Compute, responsibility for procuring and deploying the Nvidia B300 cluster. The subsidiary will oversee hardware acquisition and installation as the Dallas environment becomes operational. Meanwhile, the phased deployment provides flexibility to expand compute resources as additional systems enter production, helping the company align infrastructure delivery with customer commitments. Limestone Networks, founded in 2007, operates cloud and colocation infrastructure across multiple international markets. Its portfolio spans major US cities including Dallas, Los Angeles, Miami, New York City, Chicago, Portland, Salt Lake City, and Washington, DC, alongside facilities in Frankfurt, Paris, Singapore, and Tokyo. While the company publicly identifies the DFW1 facility at 400 S Akard Street in Dallas, it has not disclosed technical specifications or location details for DFW3.

Neocloud Strategy Continues to Expand Through Partnerships

Kidz AI entered the market in May 2026 after education technology company Classover shifted its business toward AI cloud services. Since then, the company has steadily expanded its infrastructure ecosystem through strategic partnerships designed to accelerate enterprise AI deployment. Earlier agreements included a collaboration with bare-metal GPU infrastructure provider 1Legion, followed by a partnership with Cyfuture.ai that opened access to GPU-as-a-Service, Inferencing-as-a-Service, and Tier-III certified data center operations.

Those partnerships establish the operational framework around Kidz AI’s emerging neocloud platform, while the Limestone agreement secures the physical capacity required to execute customer contracts. Consequently, the Dallas deployment represents more than another colocation deal. It signals a broader shift toward infrastructure models where enterprise demand determines capacity expansion, giving neocloud providers clearer utilization visibility as AI inference becomes one of the industry’s fastest-growing compute segments.