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 : Global Crossing - GX (formerly GBLX) -- Ignore unavailable to you. Want to Upgrade?

To: Robert Sheldon who wrote (1487)	8/14/1999 4:30:00 PM
From: Frank A. Coluccio	Read Replies (2) \| Respond to of 15615

Robert, the ROI on a DS-0 is high, relative to the equivalent contribution of bandwidth in a larger flow which is represented by an integral component of 64 kb/s. In many of the projections I've seen, assumptions are made to drive home a point that if a backbone is capable of 400 Gb/s, say, then it must be equivalent to 5,160,960 voice grade services, each operating at the DS-0 level, or each one operating at 64 kb/s. Derivation: There are 24 DS-0s per T1 There are 28 T1s per T3 There are 192 T3s per OC-192 (10 Gb/s) There are 40 OC-192s, OR 7,680 T3s per 400 Gb/s flow. It is assumed that these OC-192s are all running over the same optical strand per each direction, using different wavelengths. That being the case, then there are 40 wavelengths carrying 5,160,960 possible DS-0s per 400 Gb/s flow. Some of these arrangements use multiple fibers, simultaneously space division multiplexing while also wavelength division multiplexing. 24 * 28 * 192 * 40 = 5,160,960 DS-0s Rather arbitrarily, 14 DS-0s break even cost-wise with a T1. Eight to twelve T1s break even with a T3. 90 to 125 T3s [perhaps fewer than this range] break even with a containerized OC-192. Twenty to thirty discrete OC-192s break even with a 40-lambda 400 Gb/s flow. Like I said, these are not actuals, but in some cases very close to actuals, depending on the venue. For example, a T1 between NY and Houston may cost $5,000, and a T3 could cost $40,000. Those are Big 3 averages. Or, should I say that those were Big 3 actuals until some of the newer fiber backbone carriers [you know who they are] began offering comparable services. In any event, here it becomes clear that my break even for a T3 between NY and HOU would be eight (8) T1s. Which is to say, if I have to buy nine (9) T1s, I may as well go with a full T3 for the same cost. Everything after that [from the 9th to the 28th T1s] is gravy. [[It should interest some here to know that the newer fiber baron carriers are now selling bandwidth at a 25% to 35% discount to the T3 rates I've used here.]] If you use these assumptions, and go back and run the numbers again, I would be eager to know how they compare to your original findings. And also keep in mind that the potential of filling a pipe with so many DS-0s does not grant one the right to project the fulfillment of same for comparative purposes, for a variety of reasons. Not the least of which is the added financial and logistic burdens incurred when breaking down large flows into smaller ones (like T1s and DS0s) at each point along the way. If a FON or a T is already selling those DS-0s that means that they are already filling up their pipes at the lower denomination (DS-0) which yields a higher unit level ROI than the equivalent unbroken down unit of bandwidth (albeit, costing more to provision to the end user, at the same time, but from an international historical- and cost-based perspective, yields a much higher return than the plain vanilla bits contained within the larger, and usually aggregated, data flows). The points which I have raised above are not intended to diminish in any way the robustness, or the projected magnitudes which GBLX and others will be capable of delivering once they are fired up and selling their services. Instead, I was pointing out some of the gotchas that exist in the types of analyses which are being discussed here. Also, and in some ways ironically, a better argument might be made in GBLX's favor by demonstrating that they actually do not wish to sell DS-0s, in the first place, but instead would rather sell the massive flows to other carriers who would in turn use them to their best advantage... increasingly away from the DS-0 regime. Regards, Frank Coluccio