Bucky,
First let me say that I've enjoyed your posts. You obviously have a good understanding of the dynamics both from a corporate and technology perspective in the networking market. If I may be so bold however I'd like to add a slightly different spin. Before I do though I will say that your perspective is one that is supported by many industry pundits....but we'd be remiss if we didn't articulate the position of "the other side".
But the problem with routers and IP technology is it is connectionless, and cannot discriminate among different types of network traffic.
Well true, it is connectionless. This limitation slows traffic and requires huge routing tables which increases complexity in the network. However, IP is a higher layer protocol and therefore can discriminate among different types of traffic far better than ATM. The issue is latency. IP is superior at prioriization and at identifing flows. However since packet sizes vary greatly and since routers must check every frame it's a slow technology making it less adapatable to real time voice and video.
Without major modifications it can't prioritize voice traffic or give certain priority to a particular customer who pays a little extra money.
Correct that IP can not, without modifications, support real time traffic. However IP switching, RSVP, IP precedence bit setting, etc. etc. are gunning to be just this type of modification. Many believe that since IP is the predominant architecture AT THE EDGE (have to qualify that for my ASND friends) the most cost effective way to deploy multiservice real time environments is to adapt the technology that's already there rather than ripping out an installed infrastructure. That's what companies like Csco and Bay are attempting to do.
Bucky..the remainder of your post was excellent! I agree 100% with your assesment of LU, NN, and CSCO and their ability to pull off an ATM/IP combo technology. I'd have to add BAY..and the reason I would is because I believe the IP switching could indeed be this next technology. As IP vendors scramble to make IP more real time (like ATM), ATM vendors are scrambling to make their products more application sensitive (like IP). The big players LU, NN, NT, and CSCO all have their own biases. LU, NT, and NN all are fairly well set on layer 2 technology but NN may need to bolster this a bit with a strong ATM/FR company... I don't thing they have any designs on ASND however. Layer 3, where CSCO is strong is where these other big guns are weak and I expect they'll come looking for companies like BAY to add that technology. In the meantime CSCO has recently showed signs that they are serious about multiservice...they may want to add more voice techology...perhaps partnering or buying a telecom manufacturer. However, since margins are very low in that business I would expect a partership before I'd expect a buy. JMO..
Thanks again Bucky for the great thoughts!
OG |
