Relativity Networks Pairs $22 Million Raise With $40 Million Hyperscaler Fiber Order

A $22 million SAFE financing has arrived alongside a $40 million hyperscaler order for Relativity Networks, Data Center Dynamics reported, turning the hollow-core fiber supplier’s AI data-center pitch into a funded deployment test.

The Orlando company is trying to solve a physical infrastructure problem for AI operators: power is not always available where tightly clustered GPU capacity is needed.

ChronoCore, its hollow-core fiber system, sits inside that inter-site design so operators can spread compute across geography without treating every data centre as an isolated island.

Financing Arrives With A Hyperscaler Order

The SAFE investment was more than double Relativity’s target.

New investors in the round included Rhapsody Venture Partners, Bell Ventures Inc. and Faster Than Glass LLC, giving the company financing at the same time as a commercial order from an unnamed “leading hyperscaler.”

That customer recently tested ChronoCore between two data centres.

The follow-on order matters because hollow-core fiber has often been framed as a performance upgrade; here, the procurement is tied to an AI infrastructure buyer that has already put the technology across an inter-site link.

Jason Eichenholz, Relativity’s founder and chief executive, said AI is scaling across geography rather than only inside one facility.

His argument puts fiber latency in the same planning conversation as GPU supply, power access and data-center siting.

Hollow-Core Fiber Targets The Power-Siting Constraint

Traditional solid-core fiber pushes light through glass.

Hollow-core fiber instead uses an air-filled center channel surrounded by glass structures, changing how light moves across routes where timing is critical.

The distance claim is central to Relativity’s commercial case.

Conventional fiber has typically limited data centres to about 60 kilometers, or 37 miles, from power providers or from one another when latency is the constraint.

Hollow-core fiber extends that reach to 90 kilometers, or 56 miles, creating more room to place AI capacity near available electricity while keeping distributed sites inside a tighter latency budget.

For hyperscalers, that changes the network from a back-office connection into part of the compute architecture.

A low-latency route can help distributed GPU clusters behave more like one system when land, grid access or permitting pushes new capacity away from an existing campus.

Prysmian Partnership Moves The Cable Toward Deployment

Relativity’s manufacturing path runs through Prysmian on both sides of the Atlantic.

The fiber is made in Eindhoven, while the cable assembly work takes place in Claremont, North Carolina; Orlando remains the base for couplers, characterization work and installer training.

The partnership recently produced a 24-fiber hollow-core cable in a 10-millimeter format, described as the company’s highest-density cable to date.

Dura-Line testing installed the cable in standard microducts, a practical requirement for operators that need new optical links to work inside existing construction and conduit methods.

That production detail matters because AI campuses are constrained by more than transceiver speed.

A fiber route has to survive normal installation practice, fit available ducts and support service teams that can characterize the link after construction.

The Prysmian and Dura-Line work gives the order a manufacturing and field-readiness context rather than leaving it as a laboratory latency claim.

The remaining commercial proof is not whether hollow-core fiber can improve latency in principle.

Relativity now has to show that ChronoCore can move from partner validation and one unnamed hyperscaler order into repeatable deployments across the data-center routes where power access, construction timing and AI workload placement collide.