Moving 2.5GBASE-T and 5GBASE-T Toward Standardization
By Joseph Byrne, NBASE-T Alliance Board Member and
Senior Manager, Strategic Marketing, Freescale Digital Networking Group
The quest to standardize an economical upgrade to Gigabit Ethernet that pushes speeds up to 5Gbps using existing cabling made significant strides at the recent IEEE 802 interim session where individuals from several NBASE-T Alliance members participated. During the session, the initial standardization group, dubbed the “Next-Generation Enterprise Access BASE-T PHY Study Group,” (see link) successfully developed the key items needed to advance the standardization process from the study-group phase to the task-force phase. These items included developing the objectives, Project Authorization Request (PAR), and the Criteria for Standardization Development (CSD).
NBASE-T Alliance Chairman Peter Jones spearheaded the debate about the key objectives of rate, reach, and cabling. The Alliance believes that both data rates are needed to provide a cost-effective roadmap forward from 1Gbps. The two data rates upgrade network speeds using customers’ current installed base of cable, an economical alternative to deploying new cable to support 10Gbps.
Some Study Group members expressed concern about whether both 2.5Gbps and 5Gbps rates were feasible at the conventional 100m BASE-T distance, with particular concern about the 5Gbps rate over Cat5e cable. Technical analysis and test results presented at the meeting showed that 5Gbps over Cat5e is indeed feasible. Group members concluded that although new installations should use better cable, serving the installed base and getting customers on an upgrade path is paramount. The group, therefore, formally set the definition of 2.5GBASE-T and 5GBASE-T PHYs as an objective, including support for Cat5e cable at runs up to 100m.
A secondary action taken was to set up an ad hoc group to look at main use cases for the new PHYs. Although the use cases would not be exhaustive, they would be broad enough that the group would be able to develop targets grounded in real-world situations for particular PHY parameters. The use cases will shed further light on the 5Gbps over Cat5e issue and the characteristics of enterprise environments (which have different electromagnetic noise) targeted by the new PHYs.
The CSD is part of the IEEE process. The CSD document contains basic facts about the standard in development, including economic and technical feasibility. Incorporating lessons learned in developing 10GBASE-T and empirical results from an implementation by NBASE-T Alliance member Aquantia, the PHYs are clearly feasible and the study group easily reached consensus on the CSD.
Also adopted with little fanfare following the debate about the objectives, the PAR formally specifies the scope of the standard and forecasts the length of time until the initial sponsor ballot. Containing a higher level scope than the objectives, the PAR states the standard will address topics such as MAC parameters, PHY specifications, and management objects for Ethernet frames at 2.5Gbps and 5Gbps over twisted-pair cable. The proposed timeline to sponsor ballot is January 2017.
Although two years sounds like a long time, actual standard development can happen much sooner. Having agreed to objectives and submitted the CSD and PAR, the study group can request to advance to the task-force phase at the March plenary meeting. This would allow the first task-force meeting to take place in May. This meeting will also be the first where the group replaces the study group’s eight-word mouthful of a name with the far shorter (but hardly mellifluous) designation, 802.3bz. Congratulations to the study group for its excellent work, with a special thanks to its acting chairman David Chalupsky of NBASE-T member company Intel for his leadership.