Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Silkroad -- Ignore unavailable to you. Want to Upgrade?

To: George T. Santamaria who wrote (300)	3/9/1999 2:34:00 AM
From: ahhaha	Read Replies (2) \| Respond to of 626

I mean 10 terabits/second, more than 10,000 OC-12 serial encapsulated channels dumped onto one fiber strand which is about the limit of modern fiber. The constraint is the fiber carrying capacity. The cost saving comes from one of the SR press releases. I believe I could probably cook up a hand-waving rationalization for it. It is send, receive, management, equipment and other overhead. I was hoping the thread would like to discuss the technical details of the apparatus so far laid out. This is the crux of the matter.

To: George T. Santamaria who wrote (300)	3/9/1999 8:13:00 AM
From: Frank A. Coluccio	Respond to of 626

I find your last question extremely interesting. Although I don't think that AHhaha had this in mind, I believe that the total cost of ownership for all necessary infrastructure should be the key here. Regardless of whether we're talking about 500 Gb/s or 1 or 10 Tb/s. What supporting external-to-the-box infrastructure is necessary to administer the I/Os? Internal to the box, ditto. Can the device best utilize dispersion shifted fiber, singlemode, LEAF, what? How about leveraging existing route amplifier huts along the rights of way? What are the spacing intervals between amplifiers required in the SR model, in other words? What kind of loss budget does it manifest, end to end? How susceptible is it to backscatter effects, what's the tolerance to nonlinear distortion/SNR, etc.? How difficult will it be to engineer for different topologies and distances? I'll refrain from posting the entire laundry list for now. But the foregoing are among the factors which must be brought into the calculus before we begin to answer the questions of overall efficiencies. We're nowhere near there, yet. Not by a long shot. If we "are" going to make comparisons, for the purposes of assessing efficiencies, the only other model for these data rates against which to compare is DWDM, thus far. In DWDM each lower order stream requires its own laser-driven ensemble which includes the mechanics of optical components and ingress & egress fabrics. How do the inside AND outside factors compare between the two models? These overall considerations will play a large role in determining efficiencies, not just the costs of the indigenous components of each box, themselves. Like I've stated, I don't think we're at that stage of analysis here yet. Due to time and focus constraints, I've yet to characterize this thing in my mind's eye as a coherent entity in the larger picture, much less the unit itself. But we're getting there, slowly.