800G

800G refers to data transmission at 800 gigabits per second. In Ethernet and optical networks, it denotes a higher-speed tier following 400G, designed to provide greater aggregate bandwidth for data centers, cloud networks, and long-haul backbones.

In practice, 800G is achieved by aggregating multiple lanes into a single interface, such as eight 100G lanes or four 200G lanes. Each lane uses high-speed modulation and error correction to maximize efficiency and reliability, with PAM4 commonly used on shorter links for improved spectral efficiency and forward error correction to manage noise and distortions. Reach requirements vary by application.

The hardware ecosystem includes multi-lane pluggable transceivers in data-center form factors (for example, QSFP-DD) and server-to-switch line cards that support parallel optics or coherent-like signaling for longer reach. The approach emphasizes compatibility with existing 400G and 100G ecosystems while enabling higher port densities and lower power per bit through advanced optics and packaging.

Standards development and market status: 800G-related work is under discussion within Ethernet standards bodies and optical networking forums. Several vendors have demonstrated 800G demonstrators and begun early deployments in select environments. As with previous generations, successful adoption depends on cost, power efficiency, interoperability, and the availability of compatible fiber infrastructure and transceiver supply chains.

While 800G will not replace 400G overnight, it is positioned as a scalable option for future data-center and service-provider networks, bridging toward even higher speeds and denser interconnects. Related topics include 400G Ethernet, 100G Ethernet, PAM4 modulation, and optical fiber.