That's the update from John D'Ambrosia, chair of the IEEE 803.2ba committee. "We made a lot of forward progress at the last meeting," D'Ambrosia said.

"I was very excited to see the group moving forward because it gives me a little room," he added. "I still have people breathing down my neck for a schedule."

Currently, D'Ambrosia estimates 802.3ba could approve a preliminary work group draft as early as March and a more final sponsor draft in November 2009. As usual with standards-in-progress, when engineers decide the drafts are solid enough to implement products "depends on when people feel they can take a risk," he said.

The group is trying to define a common set of standards for both 40 and 100G Ethernet. However, it will have to define specifics particular to different configurations of target data rates and distances over specific media types.

For example, the group currently has two proposals for running 40G over 10 kilometers, a serial approach and another using four 10G lanes. "This is an interesting debate for us to have," said D'Ambrosia.

Indeed, the group has yet to face many decisions about the extent it uses multiple lanes of existing 10G Ethernet or pushes for a move to 20, 25, 40 or even 100G serial approaches over any given copper or optical media. The group also will consider backplane standards.

"People have worked on 10G but not using it in multiple lanes," D'Ambrosia said. So eventually engineers may be "looking at a lot of work aggregating 10G lanes," he said.

Designers in optical media will need to explore the possibilities for four 25G lanes supporting 10-40 kilometers. Its likely multimode fibre designs will be asked to support ten 10G lanes for some runs of 100G Ethernet, D'Ambrosia said.

So far, the .3ba group has defined a baseline for a physical coding sublayer that provides a lane-binding mechanism for both 40 and 100G rates. It has also adopted a media independent interface proposal that provides a logical definition but does not define the electrical details.

Several other baseline proposals are pending. The group has yet to define an attachment unit interface that would provide a chip-to-chip link at 40 and 100G, similar to the XAUI standard in use for 10G today. It also needs to define a physical media attachment layer which acts as a multiplexing block.

Attendance at the .3ba meetings continues to grow with about 20 additional people at the last meeting. D'Ambrosia said he expects as many as 200 will attend the next plenary session.