Why AI Needs Certified Carrier Ethernet

As organizations scale their AI operations, most find they must meld distributed compute and data resources together. They may train on one set of data, perform inference on another set in a different location, update models with yet other datasets created on the fly, and more. Typically, the elements of such operations are geographically scattered, and thus the networking of these elements becomes a critical factor for success. Enter Carrier Ethernet.

Carrier Ethernet is a standardized, carrier-grade form of Ethernet designed to deliver predictable, scalable, and service-assured connectivity over wide-area networks. Defined and governed by industry bodies such as Mplify (formerly MEF), Carrier Ethernet extends familiar Ethernet technology with attributes required for service provider and enterprise use, including deterministic performance, traffic engineering, resiliency, multi-class services, and formal service level agreements (SLAs).

Unlike best-effort Ethernet, Carrier Ethernet has been in use for more than two decades and is built to operate across metro, regional, and global networks, enabling consistent connectivity between data centers, clouds, and enterprise sites while supporting automation, orchestration, and interoperability across multiple providers.

See also: Why Modern AI Needs NaaS

Why AI Needs Carrier Ethernet

Carrier Ethernet is increasingly used for AI because AI workloads place far more stringent demands on network behavior than traditional enterprise applications. Distributed AI training and real-time inference require low and predictable latency, high bandwidth, lossless or near-lossless transport, and deterministic performance across data center, edge, and cloud environments. Carrier Ethernet provides the operational assurance, automation, and scalability needed to reliably interconnect GPU clusters, edge inference platforms, and AI data pipelines. As enterprises deploy AI across hybrid and multi-domain architectures, Carrier Ethernet provides a proven, standards-based transport foundation that meets today’s enterprise requirements and the emerging needs of AI-driven infrastructure.

Last fall, at Mplify’s GNE 2025 conference, Mplify CTO Pascal Menezes put the network aspect of AI into perspective with his opening remarks. “The network becomes critically important. It has to have performance guarantees and deliver quality and security,” said Menezes. “It also has to be dynamic, fluid on demand, programmable, and automated.”

He noted that Mplify had noticed a surge in demand for Carrier Ethernet certifications from providers of all types. They needed the certification because enterprise and service provider customers were adding certification requirements to their RFPs for transport and connectivity services for their AI efforts.

The surge was quite unexpected. Carrier Ethernet has been in use for over two decades and is widely deployed and utilized. Mplify, under its previous name MEF, was instrumental in Carrier Ethernet’s success due to its frameworks, Lifecycle Service Orchestration (LSO) APIs, and certifications.

To support the new demand for certification, Mplify redefined Carrier Ethernet certification for the AI era. Specifically, Mplify’s Carrier Ethernet certification now defines two complementary profiles, Carrier Ethernet for Business and Carrier Ethernet for AI, that validate readiness for both enterprise and AI-driven services.

Doubling Down on Carrier Ethernet Certification

This week, Mplify announced that it has restructured and expanded its Carrier Ethernet certification portfolio to better align with the accelerating demands of enterprise networking and AI-era infrastructure. Building on the established MEF 3.0 Carrier Ethernet framework, the organization has launched Mplify Carrier Ethernet for Business as the direct successor to the MEF 3.0 certification while introducing Mplify Carrier Ethernet for AI. The latter is a new profile designed specifically to validate network performance for demanding, distributed AI and agentic AI workloads.

The updated certifications provide a seamless transition path for existing Mplify-certified members and incorporate an annual maintenance and optional retesting regime to keep credentials current and market-relevant.

From an AI infrastructure perspective, the Carrier Ethernet for AI certification serves as an important industry benchmark for independent, third-party validation of transport capabilities critical to AI workloads, including low latency, high throughput, and multi-domain interoperability across distributed environments. By preserving the technical rigor and global recognition of legacy Carrier Ethernet standards while expanding into AI-optimized transport assurance, Mplify’s enhanced certification suite aims to bolster enterprise confidence in network readiness for both traditional business connectivity and next-generation AI service delivery.

A Final Word

As AI deployments increase and scale, organizations will need AI-ready infrastructure capabilities. The new Mplify Carrier Ethernet certifications provide independent, third-party validation of network performance for demanding AI workloads. 

Aligned with current Carrier Ethernet standards, the certification supports interoperability across the AI ecosystem, enabling service and technology providers to demonstrate proven capabilities for reliable, flexible, and efficient connectivity. These capabilities are critical to coordinating AI models and peripheral devices across distributed infrastructure.