SI
SI
discoversearch

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 : LAST MILE TECHNOLOGIES - Let's Discuss Them Here

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Warren Gates who started this subject7/18/2000 10:23:19 AM
From: Stephen L  Read Replies (1) of 12823
 
Circuits vs MPLS

The gauntlet has been thrown down by futurist G Gilder. The debate has raged on the Ciena thread. And Start-up de Jour's are grappling with which way to go and for how long. Perhaps we should revive the networking topology debate here. I will play devil's advocate. In the long run (aside from all being dead) it is very likely that a new version of circuit switching involving lightpaths will replace IP and MPLS routing in the backbond and Metro loops. The light links will be protocol and transmission speed neutral. The super-abundance of fiber and DWDM channels will permit enough channels to each users to facilitate multicast apps and robust topologies (N-dimensional Toruses etc) solving QoS issues. Switches will be all optic (and not necessarily that fast to change) but activated by an electronic provisioning network that will set up and tear down lightpaths at the beginning and end of a session and remain dormant during transmission. The rapidly provisionable circuit seems to exploit the sequential correlation or relationship of data streams that IP chops up (its slow but only has to make up its mind once). IP may remain at the edge and part way into the access network for awhile but eventually it switches all the way....

For an interesting perspective of switching see

lightreading.com

Now that my coffee has kicked in, yes I acknowledge the realities of legacy system (SONET migrating to IP over glass), the limited supply and high cost of optical switches and transciever (Bell Labs is working on this), the obstacles of different fiber typs, different DWDM tranmission technologies, different data speads, protocals, and the scarcity of "good" fiber. All of these point to the near term (2 to 5 year) need for an intelligent network and MPLS or its Lambda analogue seem to be very promising. Still one article pointed out that the crash you heard at the office today was the Terabit router causing the floorboards to collapse.

For a good perspective of Avici's MPLS (this is not an endorsement) see:

lightreading.com

Any comment and ideas would be appreciated,
Steve
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext