Axe Compute signed a 36-month agreement valued at approximately $260 million to deliver a dedicated cluster of 2,304 NVIDIA B300 GPUs alongside AI-focused high-speed storage. The cluster will be housed in a single U.S. Tier 3 data center and is secured via deposit, prepayment, and monthly take-or-pay payments.

This marks the largest enterprise engagement in Axe Compute’s history, providing the company with long-dated income visibility and establishing a new benchmark for dedicated AI infrastructure contracts. The deal underscores a shift toward customer-specified deployments, as enterprises demand dedicated capacity rather than shared cloud resources.

The deployment is supported by 4.8 megawatts of N+1 redundant power capacity and enterprise-grade service levels, with targeted start in Q3 2026. Customers gain predictable monthly pricing, delivery guarantees, and optional value-added services like dedicated local loops to support large-scale model training, fine-tuning, inference, and data processing workloads.