Relativity Networks Raises $22M to Bring Faster Hollow-Core Fiber to Data Centers

Relativity Networks Raises $22M to Bring Faster Hollow-Core Fiber to Data Centers

Relativity Networks has raised $22 million in SAFE note funding as the company works to bring hollow-core fiber technology to data centers and help infrastructure developers address growing networking and geographic constraints caused by the rapid expansion of AI computing.

The funding was provided by investors including Rhapsody Venture Partners, Bell Ventures Inc. and Faster Than Glass LLC. Alongside the financing, Relativity Networks also secured a $40 million follow-on order from an unnamed leading hyperscaler.

The company is developing hollow-core fiber, an alternative to conventional fiber-optic technology that uses a hollow chamber at the center of the cable instead of transmitting light through solid glass.

According to Relativity Networks, hollow-core fiber can transmit data up to 50% faster than conventional fiber. The technology works by allowing light to travel through a vacuum-like hollow core, bringing transmission speeds closer to the theoretical limit of light.

The latency difference becomes particularly important over longer distances. Relativity Networks CEO Jason Eichenholz estimates that a signal takes roughly five microseconds to travel one kilometer through conventional fiber. With hollow-core fiber, that figure can fall to around 3.5 microseconds.

While a difference of microseconds may appear small, the impact becomes more significant as AI infrastructure grows increasingly distributed.

Early AI computing systems often concentrated GPUs within a single rack or tightly connected cluster, making the latency introduced by fiber relatively easy to overlook. Modern AI infrastructure, however, is increasingly spread across large data center campuses containing multiple buildings and, in some cases, across separate campuses.

This geographic expansion is being driven in part by the enormous power requirements of AI workloads. Data center developers are increasingly forced to locate computing infrastructure where electricity and other resources are available, rather than concentrating every component in a single location.

Relativity Networks believes faster fiber could make it easier to connect these geographically separated facilities while maintaining the low latency required for AI workloads.

The company sees a particular opportunity in multi-campus deployments, where separate data centers can potentially operate as a coordinated computing system. By reducing network latency, hollow-core fiber could allow developers to place computing infrastructure farther apart without creating the same performance penalties associated with conventional fiber.

The company argues that cutting transmission latency by roughly 50% could allow AI infrastructure operators to span significantly greater distances before network latency becomes a limiting factor.

The opportunity comes as investment in data center infrastructure accelerates around the AI boom. Developers are expected to spend trillions of dollars on data center infrastructure through the end of the decade, while constraints around electricity availability, permitting and location continue to influence where new facilities can be built.

Relativity Networks is positioning hollow-core fiber as a potential networking layer for this increasingly distributed infrastructure. Rather than requiring all computing capacity to be concentrated in one location, faster interconnects could give operators more flexibility in how they distribute GPUs and other AI infrastructure.

The company describes this shift as a progression in AI infrastructure optimization. The first phase focused heavily on increasing computing power through GPUs, while the next focused on improving networking within data centers. Relativity Networks believes the next phase could involve optimizing the physical geography of computing infrastructure itself.

The latest funding and hyperscaler order give Relativity Networks additional capital and commercial momentum as it works to bring hollow-core fiber into wider data center deployments.

The challenge will be turning the technology's latency advantages into a broadly deployable infrastructure solution as AI data centers continue to scale across larger physical footprints.