The IEEE officially announced today the formation of a study group to tackle 400 Gigabit Ethernet. IEEE members approved the formation of the study group two weeks ago on the strength of an 87-0 vote (see “Efforts toward 400 Gigabit Ethernet begin”). According to the study group’s new chair, John D’Ambrosia of Dell, this consensus is an indication that the industry recognizes the need for a higher data rate three years after the approval of IEEE 802.3ba 40 and 100 Gigabit Ethernet, as well as the fact that 400 Gbps is the right next step. But that doesn’t mean the study group’s members don’t have its work cut out for it.
In a conversation last week, D’Ambrosia said that it took less time to form a consensus around the need for the new 400-Gbps specifications than it did for what became IEEE 802.3ba. The findings of the IEEE 802.3 Ethernet Bandwidth Assessment report, which indicated that networks will need to support 1 Tbps in 2015 and 10 Tbps by 2020, undoubtedly helped the cause (see “New IEEE 802.3 group seeks consensus for next-generation Ethernet”). However, that doesn’t mean the run up to the study group’s approval lacked debate. In particular, some believed that the Ethernet community should shoot for 1 Tbps rather than settle for what they considered an interim step.
D’Ambrosia believes that the 400-Gbps target makes sense, both technically and economically. “It wasn’t a question of doing it,” he explained in discussing the need to increase speeds beyond 100 Gbps. “It was a question of doing it at the right cost.”
One reason 400 Gigabit Ethernet represented the “right cost” approach is that it will be better able to build on the work now under way for 100 Gigabit Ethernet, particularly efforts around electrical interfaces using multiple lanes of 25 Gbps. D’Ambrosia expects the standards makers will rapidly adopt a 16x25-Gbps approach for the electrical interface for 400 Gigabit Ethernet. However, he acknowledges that an interface based on 50-Gbps lanes, for “second generation” applications, likely will be discussed as well.
The study group, as its name implies, will study the necessity for a 400-Gbps version of Ethernet, create a project authorization request (PAR), and establish objectives that a working group will attempt to meet if the project is approved. As part of this process, the study group must determine if work toward 400 Gigabit Ethernet would meet the IEEE's "5 Criteria." The criteria include
- Broad market potential
- Compatibiilty with other Ethernet work
- A distinct identity versus existing Ethernet standards
- Technical feasibility
- Economic feasibility
D’Ambrosia expects it will take his study group 8 to 12 months to complete its work, before the standard potentially moves to the working group level. If one uses the time it took to complete the 802.3ba standards as a model, a 400 Gigabit Ethernet standard (which could be called 802.3bs or 802.3bt) might achieve ratification sometime in 2017.
Of course, one thing that might push that date back would be if the 400 Gigabit Ethernet effort diverges to include 1 Tbps as well, just as the 802.3ba ended up covering 40 and 100 Gbps.
“I don’t really want to go through that again,” D’Ambrosia said of the possibility.
For more information on high-speed transmission systems and suppliers, visit the Lightwave Buyer’s Guide.