[AusNOG] 100G/200G/400G coming down the pipe (so to speak)

[Paul Brooks]

On 8/03/2012 11:56 AM, Ben Dale wrote:
> On 08/03/2012, at 10:27 AM, Darren Ward (darrward) wrote:
>
>> Using 2 or 4 waves for 100G is a very very bad way to do it and is only looked at by vendors who don't have a real 100G transmission solution so they just convert the short haul multilane 'lanes' into separate channels for transmission
>>
>> 100G MH/LH/ULH DWDM is single 50GHz lambda which is the efficiency difference, one optic, one channel vs 10x10G waves and 10x10G optics which is the cost benefit Matthew mentions
>>
>> I hate mentioning anything vendor on this list and so will keep it to technology specifically so a couple of vendors use coherent optics in a way called CP-PMQDSPK for example to run a single channel 100G 50GHz wavelength. 
>>
>> Still cheaper than 10x10G despite adding the coherent DSP...
Darren - might be cheaper than 10 x 10G, but if you only want 1 or 2 channels of 10G I
suspect n x 10G still wins over a 100G muxponder for n<3 or 4 - I'm sure you know
where the crossover point is!


> Thanks Darren - that's good info to know.  Using 4-lanes on LH was my assumption based on the way it is done on some short-range transceivers, so such a beast may not even exist in the real world ; )

You might have been thinking of 40G and 100G ethernet, which is defined as 4 x 10G and
10 x 10G in a non-ITU grid. Very different beast to 40G and 10G WDM long-haul.



-- 
Paul Brooks               |         Mob +61 414 366 605
Layer 10 Advisory         |         Ph  +61 2 9402 7355
-------------------------------------------------------
Layer 10 - telecommunications strategy & network design