OK, so I have used that famous “Top Gun” line before, but why shouldn't I. It is probably the best line that defines my industry efforts, especially now. I just made a request for an official Call For Interest (CFI) on 400-Gigabit Ethernet at the IEEE 802 Plenary in March.
For those who care, it will be in Orlando, so come on down and make a vacation out of it. I am hopeful that I will have my Super Bowl moment: A successful CFI, followed by a trip to Disney World!
Back in 2006, I ran into a lot of opposition starting a next higher speed effort. This time around, everyone seems to understand that we have to do it. I guess the Bandwidth Assessment we did demonstrated the degree of the problem in networking.
There is also very little debate among those participating in the IEEE 802.3 Higher Speed Ethernet Consensus Ad Hoc that the next speed is 400 Gigabit. Still, I wonder if a “rate debate” will start here similar to ones I have seen on other blogs.
Among the threads of that debate, some have said the next-rate debate began raging before the 40G/100G standard had even been ratified. The relatively underwhelming success of 40G (OC-768) in the carrier community had probably initially biased some against making another big increase in speed. However, analysts covering the data center market see 40GE making significant headway, and I would argue its attraction is its cost.
Others say the industry should make the investment in a Terabit Ethernet standard. That's easy to say when it is someone else's money.
Let's remember that everyone needs to make a return-on-investment to justify their participation. To take this argument to an extreme, why not target 10 Terabit Ethernet instead?
This is not the Field of Dreams. Voices who whisper “build it and they will come” may be ignored. Simply put, Terabit Ethernet would take too long to develop, cost too much to buy and it would have a very small market. Ultimately, this work is about providing a higher capacity link at a lower cost per bit.
If we wait a little longer, Terabit Ethernet will be feasible, some may argue. On the other hand, many people are concerned about getting caught with no new solutions to address the bandwidth tsunami we all know is coming. The development of a new rate takes time, first for the standards initiative, then for the industry ecosystem and solutions.
As I listen to all these arguments, I think about the five criteria the IEEE 802 group uses--broad market potential, economic feasibility, technical feasibility, distinct identity, and compatibility. I say that, however, fully knowing that within a few years of ratifying a 400 Gigabit Ethernet standard, we will need to “get the band back together” to once again ratchet up the speed.
I don't have much choice. My friends tease me that I am not allowed to retire until Terabit Ethernet--or beyond!--is a reality.
400 GbE is the only number that makes sense, we will not be able to do a Terabit...and even for 400 GbE it will have to be a parallel I/O interface, a serial 400 GHz driver is not in cards in CMOS...Kris
Good luck with the IEEE efforts John. I know our standards guy Pete Anslow agrees with you that 400GbE is the logical next step, as he detailed here: http://www.ciena.com/connect/blog/400GbE-versus-Terabit-Ethernet-First-steps-toward-an-answer.html
Join our online Radio Show on Friday 11th July starting at 2:00pm Eastern, when EETimes editor of all things fun and interesting, Max Maxfield, and embedded systems expert, Jack Ganssle, will debate as to just what is, and is not, and embedded system.