To: dougjn who wrote (59 ) 6/6/1998 5:34:00 PM From: Frank A. Coluccio Respond to of 173
Doug, >what impact, if any, do you see from Sprint's announcement on Wcom?< No more or less than it has on anyone else, which IMO amounts to nothing in the short term. Sprint, if it's done anything at all with this announcement, has assisted others in opening up a Pandora's Box that has, for whatever reasons real and imagined, been closed for a long time. That being, announcing that it would use ATM as its candidate for delivery of end user services. They spoke in direct defiance to the digerati who support the conventional wisdom of IPdom, and they threw some in-your-face language directed squarely at some of Wall Street's most recent sweethearts, stating that the other fiber baron IP techs were using outdated approaches to networking and would be left in the dust as a result. Here, they were referring to LVLT, obviously, and other IP purists who have announced, no, make that vowed, to use a form of IP-only transport for all forms of service deliveries. Vowed, indeed. Look what LVLT _itself_ has done within the span of two short days of the FON announcement: Yesterday LVLT announced that it too is going with (Fore Systems) ATM gear in order to bolster its support of OC48c transport [please don't ask me why, becuase the best I can come up with I've already stated elsewhere on SI for some time, and that is that they will need to do this in order to meet end users in the access network] which I assume will be an IP over ATM over SONET schema, like FON's. So much for religious conviction. But the phrasing of their release lends credence that they will be using ATM in their backbone as well as the access part. Beats me. And the day prior to the Sprint announcement, what else was announced to lend credibility to ATM? General Data Comm announced that they are now incorporating the ATM Voice over ATM capability in their globally pervasive high-end Apex line of ATM nodal concentrators. To do this they are using no other than one of the darling protocols of the VoIP camp, the G.729a algorithm set, or 8-kbps compressed voice, which will yield an 8:1 advantage over conventional POTS DS0 channels of 64 kbps each. The use of 729a should lend favorably at some point to the melding of IP and ATM services, IMO. [Before I get any replies re G.729a, G.729a is fast becoming an alternative or option to G.723.1 for those who wish to implement it.] By itself, the Sprint announcement would have been a dud, IMO, but in the greater context of what else has been taking place in the past week w.r.t. ATM, I'm going to take a wait-and-see stance with this one. One thing I've got plenty of when evaluating elephant dynamics, is time. FWIW, Regards, Frank Coluccio