AT&T Express Waves Brings Fixed 100G and 400G Links to Data

AT&T Business introduces a new optical service designed for the growing transport needs of data centres, cloud infrastructure and edge computing. Known as AT&T Express Waves, the offer delivers dedicated wavelength connections of 100G and 400G across its long-haul fibre network, with a focus on fast activation, private paths and consistent performance.

AT&T Express Waves powers 100G and 400G data links

With a target turn-up time of 24 hours on designated metro-to-metro routes, Express Waves supports high-throughput use cases including AI data transport, cross-site replication and application-level transfers. According to AT&T, the service architecture provides private optical connectivity with built-in redundancy and fixed latency behaviour between facilities.

Julie Kutchinski, Vice President of Wireline Products at AT&T Business, describes the aim: "AT&T Express Waves is about giving customers exactly what they want – capacity, certainty and speed – right when their business needs it.

"With turn-up times in as little as 24 hours on speeds of up to 400G that’s delivered over a private, redundant fibre architecture, we provide ultra-low latency and reliability at enterprise scale so teams can push AI, cloud and edge workloads with confidence and keep momentum nationwide."

Fixed wavelengths over optical backbone

Express Waves operates on Dense Wavelength-Division Multiplexing (DWDM), a method of sending multiple optical signals on a single fibre by using different light wavelengths. Each wavelength is dedicated to a customer and offers fixed capacity, avoiding the variability common with shared internet routing.

Commercially available wavelengths are available at either 100G or 400G speeds. These options align with standard router optics and match data centre interconnect (DCI) patterns already used by many enterprises. AT&T cites a trial of a 1.6 terabits per second (Tbps) single wavelength – four times faster than the current 400G tier – showing the scaling potential of its optical network.

The trial, conducted on live traffic, indicates a possible future direction where higher per-wavelength capacity reduces the number of parallel channels needed. For operators, it means simplified network management and fewer physical connections, provided compatible transceivers and line cards are in place.

Fixed capacity wavelengths like Express Waves allow network planners to eliminate uncertainty in bandwidth availability and performance. They further support scheduled workflows, such as batch data replication or east–west application transfers within large campus or regional data environments.

Connecting the edge to long-haul fibre

Alongside the Express Waves launch, AT&T extends its AT&T Wavelength Edgeless metro architecture, a model first unveiled in April 2024. The design reduces the physical and logical distance between enterprise locations and the AT&T optical backbone, improving reach and performance for metropolitan connectivity.

With local metro configurations, AT&T now claims its fibre network reaches 460,000 buildings and 2.3 million businesses. The service targets both enterprises and wholesale operators, aiming to bring colocation campuses and user premises closer to backbone transport.

In practice, the pairing links metro access points with long-haul Express Waves, creating end-to-end optical paths from local entry points to national data hubs. These combined links offer predictable latency and are intended to improve reliability for cloud access, AI workloads and east–west data flows.

Fixed 400G connections from enterprise metro sites into the AT&T backbone help support low-latency regional workloads while ensuring clean handoffs into cloud providers. The Edgeless model serves operators building multi-region clusters or managing cross-cloud deployments, where consistency and bandwidth predictability are key.

Bryn Norton, Vice President of Global Technical Sales at data centre provider Equinix, outlines the company's involvement:

“AT&T’s unique route offerings play a pivotal role in the initiative, delivering direct, low-latency access that enhances reach, reliability and resilience for our customers.

“This upgrade reflects our ongoing commitment to delivering high-performance infrastructure that supports our evolving customer needs from AI inferencing to cloud services.”

Operational simplicity with private fibre

With Express Waves, AT&T offers fixed optical bandwidth delivered over private fibre, avoiding reliance on public internet paths. The model supports enterprise-grade latency, availability and redundancy, using AT&T’s managed backbone infrastructure.

Provisioning targets of 24 hours on selected routes can benefit data centre teams dealing with maintenance schedules, change control windows or burst traffic scenarios. For example, tasks such as model training, dataset seeding or full replication between facilities require known capacity and scheduling certainty.

Standardised 100G and 400G service tiers simplify integration into common router platforms and line cards, especially in carrier-neutral facilities or wholesale colocation environments. It reduces the need for custom configurations or last-minute engineering when capacity requirements increase.

With its fixed capacity, fast activation and national reach, Express Waves is positioned as a practical transport layer for high-volume applications including cloud migration, AI workload balancing and hybrid enterprise networks.

AT&T’s combination of Edgeless metro access and long-haul Express Waves creates an end-to-end solution for operators seeking deterministic network behaviour and scalable fibre capacity across regional and national footprints.

The 1.6Tbps trial highlights the future direction of fibre transport, though Express Waves remains anchored in standardised, interoperable service tiers designed to align with available hardware and real-world deployment requirements.

Express Waves delivers a consistent, fixed optical path across AT&T’s fibre backbone, bridging the performance demands of modern data centres with the operational needs of large-scale network infrastructure.