OFC/NFOEC 2012: 100G Progress

Pete Baldwin 12/5/2012 | 5:40:11 PM
re: OFC/NFOEC 2012: 100G Progress

Sterling's right -- exhibitors focused on 400G, and there was hardly a mention of terabit. In the debate about whether to go straight to terabit from 100G, it looks like "no" won this round.

jayja 12/5/2012 | 5:39:32 PM
re: OFC/NFOEC 2012: 100G Progress

I think a key issue is compatibility with the current network.  I think there is a sense that 400 Gb/s might be done in the existing ITU DWDM grid, or at least fit to it in some degreee, while 1 Tb/s is more likely to need to break new ground.  This would make a "pit stop" at 400G on the way to i Tb/s more likely.

jayja 12/5/2012 | 5:39:30 PM
re: OFC/NFOEC 2012: 100G Progress

I'm a little out of my league on this issue, rhr, but I think you're right.  However I've heard at a conference at least one large carrier oppose "Superchannels" and/or "Flexible Grid" because they want to adhere faithfully to the current ITU DWDM" Grid".  I understand that it is more likely that 400 Gb/s solutions will be more faithful to the current ITU DWDM channel plan, but this is less likely for 1 Tb/s.  My opinion:  this may delay 1 Tb/s, but it won't obviate it, and it does make a "pit stop" at 400 Gb/s more likely and perhaps give it more legs.  In any case, I do believe compatibility with the ITU-T DWDM grid and its associated equipment, architectures and implementation schemes must be considered when one compares the development and introduction of 400 Gb/s vs 1 Tb/s systems.

rhr 12/5/2012 | 5:39:30 PM
re: OFC/NFOEC 2012: 100G Progress Jayja,-áwhen you say a key issue is compatibility with the current network, is that the case?-áIs there any extra tricks at 1 Terabit that are not already done at 400G?-á

At 400G what vendors are doing is 16-QAM and two 50GHz carriers i.e. two channels of 200Gbps. One Terabit would require 5 such carriers and 250GHz.

There is the added value of spectral shaping using a transmit DSP but that is relevant to both 400G and 1 Terabit. And as you say, that does require a flexible grid.-á

So once vendors go to 400Gbps systems, is there any reason why they can't also then do 1 Terabit using super-channels?
Sign In