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: axial who wrote (1002)	10/11/2000 10:41:49 AM
From: justone	Read Replies (1) \| Respond to of 46821

Jim: Re: the article from Total Telecom on "360Networks scraps ATM in US$60m turnaround" This is not quite as bold as it might first seem, I think. With the current ATM switches, they are running IP over ATM over- my guess is- SONET over fiber. They are still running IP/ATM, but only the MPLS portion of IP. Maybe we should call this "thin ATM". The committees pushing IP are also looking at other ATM stacks to incorporate, but these guys are not ready to wait for that work to complete. They may be running this new network over either SONET or 'thin SONET', as well- that isn't clear to me at the moment. Until 10Gbit internet is ready, this is an intermediate position between IP/ATM/SONET/fiber and. IP/MPLS/10G/fiber. So maybe this is not so bold, and it uses some tested protocols (ATM, SONET), or parts of them, as a base. One thing confuses me. Some people claim that "ATM switches are historically less expensive then IP" which I found in a FAQ on MPLS; this article says the opposite. Are there any apples to apples hard $'s available to prove it one way or the other? I wonder what they are risking by not taking the full ATM stack? To my traffic oriented mind, assuming a "little" thing like management isn't an issue, this is still gambling that a variable length packet with MPLS adequately serves to handle the various media traffic as well as fixed cells with MPLS. Now the author of the internet, Lenard Kleinrock, in the July 2000 issue of IEEE Communications (I can't post it- you have to be a member of IEEE Communications Society to get it on line), notes that with shared media and multi-access systems, you are not only faced with queuing problems from stochastic demands, but also with the issues of allocating resources to a geographically distributed and possible mobile set of demands. Thus you can't model and simulate they system they way you could while developing the internet protocol back in the 1960's. Thus this may be more bold that it first seemed. We don't know if variable packets with "multi media multi node traffic" can share the same network without problems until we deploy it. I think this experiment should be watched. I hope they have good performance monitoring equipment. It may be more bold than it seems.