In Search of Ethernet’s Holy Grail

By John D'Ambrosia


OFC 2018 will be remembered by many as the coming out party for 400 Gb/s Ethernet. And while that would be a fair assessment of the show, in reality this was an exciting conference for all things optical. One aspect of technology development of particular interest to the Ethernet Alliance is the development of serial 100 Gb/s optical Ethernet solutions.

Introducing Ethernet Alliance’s Holy Cup and Holy Grail Challenges

Back in 2014 as the Ethernet community was reeling from the introduction of six new rates of Ethernet, 100 Gb/s Ethernet was the industry darling. At that time 100 Gb/s Ethernet was based on signaling bonded across 4 optical wavelengths (lambdas) running at 25 Gb/s each. While this was an elegant solution for the time, it was understood that the finish line for 100 Gb/s Ethernet would be reached when all signaling – electrical and optical – was running at a single lane of 100 Gb/s. The Ethernet Alliance recognized this, and challenged the Ethernet industry in two ways:

  • Challenge #1 – Ethernet Alliance’s Holy Cup of 100 GbE Lambda – The first company to publicly demonstrate a 100 GbE Single Lambda in a QSFP28 will win the Holy Cup
  • Challenge #2 – Ethernet Alliance’s Holy Grail of 100 GbE SFP – The first company to publicly demonstrate a 100GbE SFP will win the Holy Grail

The winners will be publicly recognized with their names forever immortalized on these “precious” artifacts.

Figure 1_Holy Grail (left) and Holy Cup (right)

While there were several 100 Gb/s serial lambda technology demonstrations at OFC 2018, ultimately, none were judged by the Ethernet Alliance to meet the stated requirements of the challenges. With that said, the Ethernet Alliance recognizes that the industry is close to solving these problems! We wanted to be sure to clarify the requirements.