To: Teddy who wrote (622 ) 4/12/1999 8:27:00 AM From: Frank A. Coluccio Respond to of 15615
Teddy, GG describes an attractive situation, depicting FRO's Global Centers in a good light. It's difficult for me to distinguish, however, how much of this is future, or, in fact, coming to full fruition at this time. It's an extremely aggressive plan to be putting into effect with a leading edge architecture. The emphasis on CSCO's 12000 and the obviation of telco SONET etc. is in keeping with a familiar theme of GG's, although it should be noted that the carriers themselves, including T and MCIWcom are taking this path as well, where the Internet backbone is concerned. To do otherwise at this point would require one to be Ludditious. At the same time, however, both FRO and the incumbents continue to use and expand traditional means of networking in their normal POTS and data services architectures, including Digital Cross Connects, Frame Relay switches, ATM Core devices, and other traditional modes of switching and transmission, nonetheless. The 12000s are a good pick at this time for high density core network routes. But I don't see this as a replacement of anything, only as an addition, when you think about it. It's an incremental measure, in other words, where there was nothing before. At least, that is the case where FRO is concerned. What the article fails to indicate is that the dual 2.5 Gb/s pipes are actually two SONET OC-48's operating at an aggregate of 5Gb/s over different wavelengths (I assume), which actually constitute SONET containerized bundles, despite implying to the contrary. It's fashionable these days to say that SONET is being vanquished in favor of IP, but other physical layer formats for the packaging of payloads, including IP, and network management provisions [self-healing attributes, particularly] do not exist yet with any appreciable ubiquity of reach. SONET-rated bundles continue to be the fundamental carrier unit in which many future IP over whatevers are packaged, measured and deployed. By the same token, there are a growing number of alternatives appearing, which are still not ready for prime time: Other formats actually are available, but they would reach node points on the network that are simply not prepared to accept them at this time, rendering them ineffective in a pluralistic setting. On a Tier 1 Backbone provider's internal network? That's another story. Anything is possible, then, within a single domain of control, as long as they can repackage data for hand-off outside of that domain. ------ Overall, I'm beginning to look more favorably at the FRO decision, now that I've had an opportunity to speak with some who are closer to the situation, and after exploring more of what's available from both entities in the way of news and trade discussions. The article you cited goes somewhere near what I've been told elsewhere, but I've heard some of the other sides to this story, as well. I feel that it may present a bit much for GBLX to digest all at once at this time, and presents somewhat of a distraction to the skills needed to fulfill their core competence... which is stringing silica strands around the globe. But we shall wait and see. Regards, Frank Coluccio