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 : The *NEW* Frank Coluccio Technology Forum -- Ignore unavailable to you. Want to Upgrade?

To: Stephen L who wrote (141)	6/9/2000 11:23:00 AM
From: Frank A. Coluccio	Read Replies (1) \| Respond to of 46821

Stephen, some good material to work with, thanks. "In terms of resizing the processor, it was my impression that the many colors of separate signals would create a surplus of channels to be distributed to end users. Each would be protocal neutral. The trick would be to build a... " I should have been clearer. I was referring to the extra load that a gigabit pipe (perhaps multiple gigabit pipes) would exert on the "router", itself, not the dwdm unit. "... routing or switching network that would set up an "optical circuit" from the end-user (which could be a single tenant's LAN and not a desk top) through a compatible metro loop onto a likewise compatable backbone and demuxed to the end users." That sounds like you're talking about extending the in-house [or intranet] backbone with static links over optical paths, and that's fine. That is a common application today over physical strands, and in some cases wdm, already. Any layer of abstraction above this would then enter into the realms of switching, and then routing, which would call for altering the IP [or whatever] makeup of the enterprise, perhaps even their DNS structure, if lambda switching were to replace routing as the ultimate discovery mechanism. I.e., it would entail a considerable amount of work if one wishes to alter the routing scheme from existing IP to that of virtual fiber (lambda) based switching to every end point. Think of a parallel exercise, as in what it would entail going from the present IP addressing scheme to one that is entirely based on ATM global addresses, or something similar. Unless, of course, we are going to emulate IP-over-legacy, like a lot of other early dark fiber adopters are doing. "The networks (LAN, Metro, BBone) are still far from being being all optical but the pieces are falling into place." Are you saying this in a futuristic sense, or now? If now, then where? Most optical that exists has simply been for SONET, GbE or other short haul LAN extension uses, some proprietary. And we still have issues with grid compatibility between vendors dwdms to content with. Another area that one must be concerned with lies in the fact that even where multiple sources of optical are available for tandem connections, there are disparate network management schemes unique to each vendor's platform which tend to glitch things up, too. "I don't thing that the O-e-O conversions required in most Metro loops will be that bad for performance. IMO, that would depend on the topology that you create, and how many buildings or wdm boundary crossings you plan to include in the network. For a simple point to point, I'd agree with you. If, on the other hand, you are referring to a sizable metro network with, say, twenty buildings on it, then the options you have to select from and the potential for doing things in a less than optimal manner is great. The providers are in early stages of experimentation right now, both with the technologies that are at their disposal (which are evolving rapidly) and the pricing for same. I don't want to say that no one has a clue as to what they are doing at this time, but if I were to read something like that in a trade mag, I wouldn't be overly startled by it. "With a potential abundance of lightpaths the number of channels will be more important than the speed at which they can be set up or torn down. Even if the edge remains electronic, with gig ether and faster technologies being deployed, I think that layer 2-3 and layer 4-7 switches could create an interesting feed into a backbone router and onto a dwdm light path." Yes, but again we can't assume that the router or anything else in the existing network will be up to it on day one, or even be able to scale to it, if the previous links never went beyond T1 or T3 in the past. There will be a substantial number of collateral modifications that must be made whenever one moves into near- or all- optical networking. FAC