Ethernet speed — enabled by new 50G electrical interfaces — allows the deployment of cost-effective 50G links from high performance servers to a new generation of switches and routers. Within these devices, 50G interfaces can be combined to build higher speed connections to other systems within the data center or out to the cloud.
According to the IEEE, 200G will be the next area for explosive growth in the Ethernet ecosystem due to the insatiable demand for more bandwidth in mobile devices, high-speed data center servers, internet-enabled entertainment (especially video), cloud computing and social media.
“The 200G Ethernet speed heavily leverages 100G 4x25G NRZ technology now shipping in high volume. Its 4x50G PAM4 architecture re-uses 100G optical packaging and lasers. This enables near-term cost-effective volume deployment,” said Chris Cole, vice president of Advanced Development at Finisar.
Spirent said that the new generation of switches and routers, based on 50G electrical interfaces, will present challenges for equipment vendors and those who deploy their equipment. It will start with ensuring the stability of the physical link as the traffic scales up to 200Gbps. Engineers also need to validate new set of features, functions, performance and scale.
“Traditionally, combining four electrical interface lanes has been the sweet spot, both technically and economically, for Ethernet technologies,” said Neil Holmquist, VP, Marketing Cloud and IP at Spirent. “Just as the use of four 25G lanes has been embraced for 100GbE, it is our belief that the use of four 50G lanes will be widely adopted for 200GbE.”