Ethernet’s Next Rates – 800 Gb/s and 1.6 Tb/s

Feb. 10, 2022
The P802.3df Task Force will define 800 Gb/s and 1.6 Tb/s Ethernet, while leveraging this work to develop derivative specifications for 200 Gb/s Ethernet and 400 Gb/s Ethernet.

Given the ever-expanding presence of Ethernet in today’s systems and networks, it is imperative for the IEEE 802.3 Ethernet Working Group to keep developing solutions that will deliver on the bandwidth needs of the industry. Based on the work of the 2021 IEEE 802.3 Beyond 400 Gb/s Ethernet Study Group, the IEEE P802.3df 200 Gb/s, 400 Gb/s, 800 Gb/s and 1.6 Tb/s Ethernet Task Force was formed. The P802.3df Task Force will define 800 Gb/s and 1.6 Tb/s Ethernet, while leveraging this work to develop derivative specifications for 200 Gb/s Ethernet and 400 Gb/s Ethernet.

Table 1 summarizes the physical layer objectives the Task Force will address. Based on the number of physical layer objectives, this project can be considered one of the largest IEEE 802.3 Ethernet projects in recent years.

A closer look, however, reveals that these objectives could be divided into four main tracks of development for the IEEE P802.3df Task Force.

  1. The first track is related to all the objectives shown in orange in Table 1 for 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet, where 100-Gb/s signaling per fiber in each direction or differential pair is assumed. These objectives relate to electrical interfaces or Attachment Unit Interfaces (AUI), backplanes, copper twin-axial cables, multi-mode fiber, and parallel single-mode fiber. It is anticipated that these solutions will leverage specifications defined by existing standards or work-in-progress within IEEE 802.3. This includes IEEE Std 802.3cutm-2021, as well as the standards being defined by the IEEE P802.3ck and IEEE P802.3db Task Forces.
  2. The second track is related to all the objectives shown in purple in Table 1 for 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet, where 200-Gb/s signaling per differential pair is assumed. These objectives relate to electrical interfaces or AUIs and copper twin-axial cables.
  3. The third track is related to all the objectives shown in green in Table 1 for 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet, where 200-Gb/s signaling per optical fiber or wavelength is assumed. These objectives relate to one, two, four, and eight fibers in each direction for 500 m and 2 km for all noted rates, as well as a four-wavelength WDM solution for 800 Gb/s Ethernet for 2 km.
  4. The fourth and last track is related to the 10-km and 40-km duplex single-mode fiber objectives shown in blue in Table 1 for 800 Gb/s. It is unclear at this time whether a direct detect optical or a coherent optical solution will be selected to address these objectives.

While many tend to focus on these physical layer objectives, it is the underlying architecture that will enable all these solutions that is the real priority for the IEEE P802.3df Task Force. For example, in Table 1, solutions based on 100-Gb/s signaling per lane and 200-Gb/s signaling per lane will need to co-exist and interoperate. Given the various physical layer objectives, potential implementations, as well as the various forward error correction needs of each physical layer objective, the importance of the architecture cannot be overstated. Furthermore, new optical packaging approaches such as co-packaged and near-package optics increase the number of implementations that must be considered for the development of the underlying architecture.

Conclusion

The IEEE P802.3df project represents one of the largest projects in recent history for the IEEE 802.3 Ethernet Working Group. It will address four Ethernet rates, 26 physical layer specifications, development of 200-Gb/s electrical and optical signaling, and possible development of 800-Gb/s coherent signaling. It will expand existing four-lane solutions based on 100 Gb/s to eight lanes to achieve 800 Gb/s Ethernet and 16 lanes to achieve 1.6 Tb/s. Using 200-Gb/s signaling, it will create one-, two-, four-, and eight-lane solutions to address multiple specifications of electrical interfaces and physical layers specifications for 200 GbE, 400 GbE, 800 GbE, and 1.6 TbE. It will need to consider the development of solutions based on 100-Gb/s signaling and 200-Gb/s signaling and their potential co-existence. Most importantly, it will develop an underlying architecture to support all of this.

Join us on February 16, 2022 for the Lightwave Webinar, “Ethernet’s Next Rates – 800 Gb/s and 1.6 Tb/s,” where participants of the IEEE P802.3df Task Force will discuss the IEEE P802.3df project, discuss these issues, and engage in an industry discussion about its needs for 800 Gb/s and 1.6 Tb/s Ethernet.

John D’Ambrosia is chair of the IEEE P802.3df 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet Task Force and distinguished engineer at Futurewei Technologies.

About the Author

John D'Ambrosia | Chairman

John D’Ambrosia is chair of the IEEE P802.3df 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet Task Force and distinguished engineer at Futurewei Technologies.

Sponsored Recommendations

Understanding BABA and the BEAD waiver

Oct. 29, 2024
Unlock the essentials of the Broadband Equity, Access and Deployment (BEAD) program and discover how to navigate the Build America, Buy America (BABA) requirements for network...

Meeting AI and Hyperscale Bandwidth Demands: The Role of 800G Coherent Transceivers

Nov. 25, 2024
Join us as we explore the technological advancements, features, and applications of 800G coherent modules, which will enable network growth and deployment in the future. During...

The Road to 800G/1.6T in the Data Center

Oct. 31, 2024
Join us as we discuss the opportunities, challenges, and technologies enabling the realization and rapid adoption of cost-effective 800G and 1.6T+ optical connectivity solutions...

How AI is driving new thinking in the optical industry

Sept. 30, 2024
Join us for an interactive roundtable webinar highlighting the results of an Endeavor Business Media survey to identify how optical technologies can support AI workflows by balancing...