Why Ethernet Never Left: A 50-Year Networking Mystery
ATM tried to replace it. MPLS carried it. SD-WAN left it behind. Fifty years later, the answer to networking's biggest mystery was Ethernet all along.
Learn
What Is SD-Ethernet?
SD-Ethernet makes private Ethernet programmable across sites, data centers, clouds and operators over any transport.
Read the BlogSoftware-Defined Ethernet (SD-Ethernet) makes private Layer 2 connectivity programmable across customer sites, data centers, clouds and independent operators. It extends Ethernet over available transport, including fiber, internet, 5G and satellite. Paths are configured automatically, encrypted by default and monitored segment by segment between devices.
- Private Ethernet over any transport
- Paths configured automatically. No routing protocols.
- Every path encrypted and monitored
What problem does SD-Ethernet solve?
SD-Ethernet solves the complexity of extending private Ethernet beyond a single location. Ethernet was designed for local networks, so stretching it across sites means coordinating access providers, configuring network devices and troubleshooting across separate administrative domains. That work grows with every site and provider added.
SD-Ethernet automates Ethernet service setup over available transport. Operators provision connections between locations and monitor performance segment by segment from one management interface.
How does SD-Ethernet work?
SD-Ethernet connects locations through an automatically provisioned private Ethernet path. An operator selects two endpoint locations. Cloud-based software finds the best path and configures the devices at each location, without routing protocols. Those devices carry and encrypt the Ethernet traffic over the available transport and continuously measure performance between them.
-
At each location
Purpose-built devices provide an Ethernet handoff alongside existing routers and switches, without changing the core network.
-
In the cloud
Software finds the best path and configures the participating devices automatically. No routing protocols required.
-
Along every path
Traffic is encrypted between devices, with continuous measurement of latency, jitter and packet loss on each segment.
What are the key capabilities of SD-Ethernet?
Any transport
Extend private Ethernet over fiber, internet, 5G or satellite, using the connectivity already available at each location.
Programmable
Create, change and remove Ethernet paths through software or an API, so connectivity adapts as requirements change.
Automated
Select two endpoint locations. Software computes the best path and configures the participating devices automatically, without routing protocols.
End-to-end visibility
Latency, jitter and packet loss are measured continuously between devices, with segment-level visibility to help pinpoint problems.
Encrypted by default
Traffic is encrypted between devices on every path, including when it crosses shared networks.
How is SD-Ethernet different from SD-WAN, Carrier Ethernet and MPLS?
SD-Ethernet focuses on automating private Ethernet connectivity across locations and participating operators, with encryption and segment-level visibility. It complements SD-WAN, Carrier Ethernet and MPLS rather than replacing them.
| Technology | Primary role | How SD-Ethernet relates |
|---|---|---|
| SD-WAN | Connects sites and steers IP traffic based on application needs and network conditions. | SD-Ethernet extends Layer 2 connectivity and can provide transport for an SD-WAN service. |
| Carrier Ethernet | Defines standardized Layer 2 services connecting customer locations. | SD-Ethernet automates Ethernet path setup over available transport, including across participating operators. |
| MPLS | Forwards traffic using labels and supports services such as Layer 2 and Layer 3 VPNs. | SD-Ethernet can use an existing MPLS network as transport to extend Ethernet services. |
- SD-Ethernet vs. SD-WAN
- Both can steer traffic across multiple transports. SD-WAN focuses on application-aware IP connectivity, while SD-Ethernet extends private Layer 2 Ethernet services across locations and participating operators. They can work together: an operator can deliver an Ethernet connection through SD-Ethernet and run a managed SD-WAN service over it.
- SD-Ethernet vs. Carrier Ethernet
- Carrier Ethernet defines standardized Layer 2 services. SD-Ethernet provides a way to deliver those services over available transport, including beyond an operator's fiber footprint, with automated path setup, encryption and segment-level visibility.
- SD-Ethernet vs. MPLS
- SD-Ethernet doesn't replace MPLS. Operators can retain their MPLS backbone and use SD-Ethernet to reach additional sites over other available transport. This allows them to extend the Ethernet service without deploying MPLS at every new customer location.
Who uses SD-Ethernet?
Service Providers
Fiber operators, carriers and MSPs use SD-Ethernet to reach off-net customer sites and deliver private Ethernet services under their own brand. Once equipment and underlying connectivity are in place, services can be activated in minutes.
Learn more →Data Centers
Data center operators use SD-Ethernet to connect facilities, simplify tenant connectivity and extend their network presence into other locations over existing transport.
Learn more →AI Infrastructure
Neoclouds and GPU infrastructure providers use SD-Ethernet to connect distributed sites and give customers private, encrypted and monitored connectivity to GPU infrastructure.
Learn more →“The solution is remarkably simple; we plug it in and start passing traffic in minutes.”
How does MaiaEdge deliver SD-Ethernet?
MaiaEdge combines cloud-based path computation with hardware deployed at network edges, data centers and customer sites:
- Path Computation Engine (PCE): runs in the MaiaEdge cloud, computes the best path between selected endpoint locations and configures the participating PBCs automatically, without routing protocols.
- Path Border Controller (PBC): carries Layer 2 traffic over available transport, encrypts traffic between PBCs and continuously measures latency, jitter and packet loss on each segment.
Operators provision Ethernet services and monitor performance from one dashboard, without changing their core network.
SD-Ethernet FAQ
Is SD-Ethernet the same as SD-WAN?
No. SD-WAN connects sites at Layer 3 and steers IP traffic based on application needs and network conditions. SD-Ethernet extends private Layer 2 Ethernet connectivity across locations and participating operators, with automated provisioning, encryption and segment-level visibility. The two can work together: SD-Ethernet can provide the underlying connectivity for a managed SD-WAN service.
Does SD-Ethernet replace MPLS?
No. Operators can keep MPLS in their core network and use SD-Ethernet over existing transport, including MPLS, to extend private Ethernet beyond their existing MPLS footprint. That lets them retain their backbone without extending MPLS to every customer site or cloud connection.
Do I need new hardware?
Yes. MaiaEdge SD-Ethernet uses purpose-built devices called Path Border Controllers (PBCs), deployed alongside existing routers and switches at network edges, data centers, points of presence or customer sites. You don't need to replace your core routers or backbone. PBC hardware is offered as a subscription with MaiaEdge's cloud software.
Is SD-Ethernet secure?
Yes. MaiaEdge SD-Ethernet encrypts traffic between Path Border Controllers by default using AES-256-GCM IPsec, protecting it as it crosses shared or third-party networks.
Does SD-Ethernet work across different providers' networks?
Yes. SD-Ethernet can carry private Ethernet over transport supplied by multiple providers. When participating operators interconnect their SD-Ethernet deployments, they can also extend services across each other's networks while retaining control of their own infrastructure. This forms the foundation for MaiaEdge's Federated Private Networking.
How fast can a connection be activated?
Once PBCs are installed and underlying connectivity is available, operators can activate a private Ethernet path in minutes through a dashboard or API, compared with the weeks to months a traditional Ethernet circuit can take. Sites can start on existing internet access and later move to a fiber connection while retaining the same Ethernet service.
Does SD-Ethernet support Q-in-Q and VLANs?
Yes. SD-Ethernet preserves customer VLAN tags across the Ethernet path, extending existing segmentation between locations. Q-in-Q adds an outer service VLAN tag, allowing operators to keep customer traffic separate while carrying each customer's VLANs over shared infrastructure.
Is SD-Ethernet an industry standard?
No. SD-Ethernet is an architecture for programmable private Ethernet, not a formal industry standard. It builds on established Ethernet and VLAN standards, including Q-in-Q. MaiaEdge is a member of the Mplify Alliance, formerly MEF, which develops Carrier Ethernet and network service automation standards.
Read the story
