Optical fiber has been the foundation of global communications since the late 20th century. It enabled the spread of the internet, cloud services, mobile data, and digital collaboration by carrying information as light pulses across continents and under oceans. Traditional fiber optics use a solid glass core to guide light, but this medium inherently slows signals because light travels more slowly in glass than in air or vacuum. This physical limitation constrains the ultimate speed and latency that fiber networks can achieve.
Recent advances in optical physics and manufacturing have brought a new class of fiber to the forefront of telecommunications research and deployment. Known as hollow‑core optical fiber or HCF, this technology guides light primarily through an air‑filled center rather than solid glass. Because light travels faster in air than in solid silica, hollow‑core fiber offers significant reductions in signal loss and latency compared with conventional fiber. Over the past few years, sustained research efforts have advanced HCF from a theoretical curiosity to an emerging technology with real commercial deployment, particularly by prominent companies such as Microsoft in collaboration with the University of Southampton.
This article examines the science behind hollow‑core fiber, the latest performance breakthroughs, the implications for telecommunications and artificial intelligence networks, the challenges to widespread deployment, and the broader impact on global communications infrastructure.
Optical Fiber Fundamentals
Conventional optical fiber is built around a glass core surrounded by a cladding material with a slightly lower refractive index. Light entering the core reflects internally at the boundary of the core and cladding, which keeps it confined over long distances. Because silica glass has a refractive index significantly greater than air, light travels through it at an effective speed lower than it would in a vacuum or in air. Typical propagation speeds in solid‑core fiber range between approximately 180,000 and 200,000 kilometers per second. As a result, even high‑speed fiber links introduce latency that accumulates over long distances.
Latency and signal degradation are two fundamental performance considerations in network design. Signal loss, also known as attenuation, increases with distance. To maintain signal integrity in long‑haul links, optical amplifiers are installed at regular intervals to boost weakened signals. This adds cost, complexity, and energy consumption to network operations. Because of these limitations, researchers have long sought ways to reduce both latency and attenuation in optical communications.
The Emergence of Hollow‑Core Fiber
Hollow‑core fiber is an alternative architecture in which the central region is filled with air or vacuum rather than solid glass. Light propagates through this air core, experiencing much less refractive index interference than within glass. Because the refractive index of air is closer to one, the effective speed of light in an HCF can approach the speed it would have in free space. This characteristic theoretically enables lower latency and higher data transmission speed relative to conventional fiber.
Early hollow‑core fibers exhibited high signal loss and other practical challenges that limited their usefulness. Losses in early designs were significantly higher than what standard single‑mode glass fiber could achieve, making them unsuitable for long‑distance communications. However, recent breakthroughs in design and fabrication have changed that outlook dramatically.
Breakthrough Performance With DNANF Design
Advances in hollow‑core fiber technology have culminated in the development of the Double Nested Antiresonant Nodeless Fiber (DNANF) design. This innovation combines an air core with multiple concentric, ultra‑thin glass membranes that act as precise reflectors, guiding light within the hollow region with minimal interaction with glass material. These nested layers significantly reduce signal leakage, absorption, and scattering effects that previously plagued hollow‑core designs. optics.org
In a study published in Nature Photonics and confirmed in laboratory tests in 2025, researchers from the University of Southampton and Microsoft reported that a DNANF could achieve signal loss of 0.091 decibels per kilometer (dB/km) at the 1,550 nanometer wavelength, a common wavelength for optical communications. This performance surpasses the long‑standing attenuation floor of around 0.14 dB/km associated with conventional high‑purity silica fiber, a figure that remained essentially unchanged for decades.
The new HCF maintained losses below 0.2 dB/km over a broad spectral window of roughly 66 terahertz. Maintaining low attenuation across a wide frequency range is important because it allows network designers flexibility in choosing wavelengths for optimal performance, cost, and compatibility with existing optical components.
One of the most remarkable implications of these measurements is that an HCF designed with DNANF architecture can allow a light signal to travel more than 30 kilometers before its power is reduced by half, compared with only about 15 kilometers for a typical solid‑core fiber. Extending unamplified span distances can reduce the number of repeaters or amplifiers required in long‑haul links, decreasing operational costs and energy consumption.
Speed Gains and Latency Reduction
Because hollow‑core fiber guides light through air rather than glass, it offers faster propagation speeds for optical signals. Articles reporting on DNANF fiber explain that the reduction in refractive interference enables transmission speeds up to 45 percent faster than in conventional fiber. This performance improvement is crucial for latency‑sensitive applications such as cloud services, artificial intelligence inference, high‑frequency trading, and real‑time interactive applications. Phys.org+1
Lower latency can significantly impact distributed computing systems where data must traverse long distances between servers and data centers. For artificial intelligence models operating across multiple cloud regions, even a few microseconds of improved responsiveness can translate into measurable performance gains for large‑scale applications. Reducing latency also benefits online gaming, virtual collaboration, and other services that depend on fast round‑trip communication.
Industry Involvement and Commercial Deployment
Among the leading voices pushing hollow‑core fiber toward real‑world use is Microsoft. In December 2022, the company acquired Lumenisity Limited, a spin‑out from the University of Southampton Optoelectronics Research Centre that operated one of the first dedicated hollow‑core fiber manufacturing facilities in the United Kingdom. The acquisition signaled Microsoft’s intention to develop and scale advanced fiber technology for its cloud infrastructure.
According to reports, Microsoft has already deployed significant lengths of DNANF hollow‑core fiber in live network traffic. The company has publicly stated that more than 1,200 kilometers of HCF are carrying data in Azure’s network, and plans are underway to expand deployment to 15,000 kilometers over the next few years to support AI and cloud connectivity.
Microsoft’s network engineering teams have emphasized that ultra‑low‑loss hollow‑core fiber will enable faster and more reliable connections with lower energy consumption. This is especially relevant for cloud data centers, where infrastructure cost and power usage are major considerations. Since data centers span global networks, improvements to core fiber links can yield system‑wide performance benefits.
Implications for Telecommunication Networks
The potential influence of hollow‑core fiber on telecommunications extends across multiple dimensions. The most immediate effect is on latency and speed. Latency improvements reduce the time it takes for signals to traverse long distances, making data exchange more responsive for real‑time and interactive services. This can support new classes of applications, such as remote control of industrial systems, immersive virtual experiences, and distributed scientific experiments.
Reducing signal loss benefits network design by extending unamplified span distances. Fewer amplifiers along a fiber link means less capital expenditure and operational expenditure for service providers. Each amplifier adds complexity, cost, and energy usage, and minimizing their number contributes to simpler and more energy‑efficient networks.
Hollow‑core fiber also exhibits lower chromatic dispersion, meaning different wavelengths of light travel at more similar speeds. Reduced dispersion simplifies optical signal processing in transceivers and may lead to lower energy consumption for coherent transmission systems in long‑haul and metro networks.
Beyond traditional telecommunications applications, hollow‑core fiber may support emerging technologies such as quantum communication, where preservation of optical coherence is critical, or distributed sensing systems that depend on precise light propagation characteristics.
Challenges to Widespread Adoption
Despite performance gains, several technical and economic factors must be addressed before hollow‑core fiber becomes a widespread replacement for conventional fiber infrastructure.
Manufacturing complexity remains a major challenge. Hollow‑core fiber designs require micrometer‑scale precision in the construction of glass membranes and microstructures. Producing consistent, high‑quality fibers over thousands of kilometers is a non‑trivial task that demands significant investment in advanced fabrication technologies and quality control.
Integration with existing network infrastructure poses another challenge. Current optical networks are built around standards and components optimized for traditional silica fibers. Connectors, splicing equipment, and amplifier technologies are designed with solid‑core fibers in mind. Adapting these systems for hollow‑core fibers will require updated standards, tooling, and engineering practices.
Standardization efforts are essential for interoperability. Telecommunications standards bodies will need to define specifications for hollow‑core fiber performance, testing protocols, and compatibility requirements. Achieving consensus across global industry stakeholders takes time and coordinated effort.
Finally, cost considerations will influence adoption pace. Initial deployment of hollow‑core fibers may focus on high‑value segments such as large cloud providers, AI networks, and long‑haul backbone routes where performance gains justify investment. In short‑distance or access network contexts where latency demands are less stringent, conventional fiber may remain the practical choice for years.
Broader Impacts and Future Directions
Hollow‑core fiber innovation is part of a broader evolution in optical communications technology. It demonstrates how advances in materials science, photonics, and manufacturing can overcome longstanding limitations in data transmission. The recent record‑low signal loss and increased bandwidth performance suggest that hollow‑core designs will play a significant role in future network architectures.
Since light travels close to its maximum possible speed in hollow‑core fiber, network designers can reimagine latency budgets for global systems. This could influence everything from 5G and future 6G mobile networks to intercontinental submarine data links linking major population centers.
The expanded usable spectral window of hollow‑core fiber, with low attenuation across a broad range of frequencies, may also support innovations in wavelength‑division multiplexing and multi‑band optical systems. These systems can carry multiple independent channels of data on different wavelengths, greatly increasing total throughput without requiring new physical fibers.
Moreover, the same technology may support new applications in high‑power laser delivery, optical sensing, and advanced scientific instruments, where transmission of light with minimal loss over extended distances is desirable.
Conclusion
Hollow‑core optical fiber represents a significant technological advance in the field of optical communications. By guiding light through air rather than solid glass, these fibers achieve both lower signal loss and faster propagation speeds than traditional fiber optics. The collaboration between Microsoft and the University of Southampton has produced breakthroughs that push hollow‑core designs past performance thresholds once thought impractical, including record signal loss figures and demonstrable live network deployments.
While challenges remain in manufacturing scaling, standardization, and integration with existing infrastructure, hollow‑core fiber is positioned to become an important part of future telecommunications systems, especially those requiring high performance, low latency, and energy efficiency. As research and deployment continue, this technology may one day underpin core segments of the world’s data infrastructure, enabling new services and driving innovation across multiple industries.
References
University of Southampton achieves record‑low signal loss with hollow‑core fiber, Optoelectronics Research Centre News. https://www.orc.soton.ac.uk/news/7210 University of Southampton
Record‑low signal loss and increased speed via hollow‑core fiber, Phys.org. https://phys.org/news/2025-09-hollow-core-optical-fiber-transmits.html Phys.org
Microsoft‑backed team achieves lowest signal loss in optical fiber, Tom’s Hardware. https://www.tomshardware.com/tech-industry/hollow-core-fiber-research-smashes-optical-loss-record Tom’s Hardware
Broadband HCF with attenuation below 0.1 dB/km, Microsoft Research publication. https://www.microsoft.com/en-us/research/publication/broadband-optical-hollow-core-fiber-hcf-with-an-attenuation-lower-than-0-1-decibel-per-kilometer Microsoft
Telecom analysis on HCF speed and attenuation improvements, ETTelecom. https://telecom.economictimes.indiatimes.com/news/telecom-equipment/revolutionary-hollow-core-optical-fibre-boosts-data-transmission-speed-by-45/ ETTelecom.com
Optical fiber fundamentals in Wikipedia. https://en.wikipedia.org/wiki/Fiber-optic_cable Wikipedia