AI cloud provider QumulusAI has secured $124 million in customer commitments.

The company has signed two three-year contracts totaling more than $124m with "open access AI cloud" Hyperbolic and another unnamed AI inference platform.

The agreements come with $21.9m in combined upfront commitments.

QumulusAI will deploy 1,280 Nvidia Blackwell GPUs under the contract, comprising 160 Lenovo and Supermicro bare-metal servers equipped with B300 and B200 Blackwell GPUs, respectively. Cisco Nexus One will power the cluster fabric for both deployments, delivering secure, high-performance AI networking.

“AI infrastructure can no longer be built using one-size-fits-all designs,” said Mike Maniscalco, CEO of QumulusAI. “Inference workloads have very different performance and economic requirements than model training environments. By tuning infrastructure to the workload itself, we can improve utilization, reduce costs, and accelerate deployment timelines for customers operating at production scale.”

“As AI adoption expands, organizations need access to infrastructure that can be deployed quickly, scaled efficiently, and aligned to the economics of production AI,” added Maniscalco. “These agreements demonstrate the value of a more flexible approach to AI infrastructure.”

QumulusAI has not disclosed where the cluster will be deployed. The company works with colocation providers across the US, including in Atlanta, Kansas City, Philadelphia, Denver, and Brooklyn.

In February 2026, the company received approval for a modular data center in Denton, Texas. The project is planned for around four acres of land near Western Boulevard and Jim Christal Road, and will operate near the RD Wells substation. It is hoped to have around 20MW of capacity. The land is being leased from the city for around $20,000 annually for the next 10 years, as agreed in August 2024.

In April 2026, QumulusAI secured a $45m convertible note facility to support the procurement of GPUs and data center infrastructure.