Steve, I've appreciated each of your upstream messages and found them consistent with your usual level of reporting clarity and discretion. IMO, very few posters put their back into detail the way you do on SI. For that, I say Bravo, and Thanks!
As for why SR hasn't placed their wares up for test in an independent national lab, I could offer only that I suspect that there are no established test beds which have been parametrically defined to examine this model yet. Coming to such a consensus for defining parameters and measurement principles, for what could amount to a one time view, could be a sizable undertaking, requiring considerable retrofitting to existing test beds and their underlying assumptions, first.
Most optical component benchmarks for DWDM are defined in terms of straight time and frequency domain metrics, and the usual electronic analogies now found in photonics with regard to non-linearity, jitter, distortion, channel crosstalk, etc. In this sense, even the independent labs accustomed to dwdm performance grading may find SR to be disruptive, to borrow from a growing, and now-popular, if not slightly over-used, concept. I say overused because at some point disruption becomes not only highly cost effective in the more global context (omitting, of course, any consideration for the impact on previous vendors' fortunes), but too commonplace, at the same time, to even make mention of it. We're getting there in this space, I believe.
Incidentally, and for what it's worth, the fact that the VC partner didn't outright disqualify or bash SR's approach, allowing that there was some chance of merit, could be described as a glowing review by some. An outright panning, OTOH, would have had a much more catastrophic impact, by far.
Of course, when the type of review (such as you described coming from that individual) results in a kind of "we'll leave the door open just in case" statement, it could also be viewed as a sign of hedging against the unknown, when in fact there was no apparent proof of any significant merit, to begin with. No one likes to be proved absolutely wrong over time, and this is a way out. 'No comment' wouldn't fly hardly as well, even if they had no clue about that which they were examining.
Re: LU's Martin, I would be owing to you if you could find the time to expand on the following, if you would:
"Lucent's CTO Bob Martin (who is a BIG Clayton Christensen fan - will this
arbitrage the red/blue zone differential as "Lucents" now look to be involved in
disruption?) is talking about 50 THz (on 1K lambdas) on a single fiber with 2:1
single channel efficiency (vs LU's BLAST program which has achieved MUCH
higher - up to 1000x - channel efficiency by going beyond the Shannon single-channel approach) by 2005. Martin claims that 400Gbps (over 80 lambdas) is available and DEPLOYED today."
My interest stems from some recent examination of schemes which would packetize optical bundles, based on colors possessing narrow line widths. TIA.
Regards, Frank Coluccio
|
