Moonshot Energy, QumulusAI (QAI Moon), and Connected Nation Internet Exchange Points (IXP.us) collaborated on a nationwide AI computing and Internet exchange (IX) platform connecting US universities and municipal networks.

The so-called AI IXP platform aims to deliver AI inferencing closer to networks and end users.

The partnership saw Moonshot Energy deliver services such as electrical systems, modular power, and cooling, while QumulusAI provided the GPU orchestration framework, workload delivery platform, and commercial model optimized for inference-intensive, latency-sensitive AI workloads.

IXP.us, meanwhile, delivered the carrier-neutral interconnection foundation, with the initiative launching with 25 pilot sites. The venture plans to scale to 125 locations across US university research campuses and municipalities, integrating IXP facilities with modular AI pods, which are blocks of compute infrastructure purpose-built to run AI workloads.

Each market deployment begins with around 2,000 kW of module capacity to support adjustable application-driven GPU configurations tailored to customer requirements and scalable AI workload growth.

QAI Moon claimed each pod deployment was designed to function as a network-dense, low-latency, optimized inference platform, architected to support its GPU-as-a-service delivery model. These require dual, geographically diverse 400 Gbps (400G) IP transit connections from four independent ISPs, complemented by redundant 400G IX ports on the DE-CIXaaS switch at each IXP.us facility.

The architecture maintains direct adjacency to high-count dark fiber links connecting IXP interconnection infrastructure with modular AI compute systems, minimizing latency and maximizing throughput.

The first QAI Moon deployment is due by July 2026, at the IXP.us “alpha site,” located on the Wichita State University campus, with expansion planned across other IXP.us markets currently in development.

“This partnership represents the physical convergence of power, compute, and interconnection at the exact point where AI demand is moving,” Moonshot CEO Ethan Ellenberg explained in a statement. “By pairing Moonshot’s modular electrical and AI infrastructure with the IXP.us carrier-neutral interconnection model and QumulusAI’s GPU platforms, we are creating a repeatable national architecture that delivers ultra-low-latency AI without the constraints of hyperscale data centers.”