Market Leader Programs
5G Transport - A 2023 Heavy Reading Survey
2023 Open RAN Operator Survey
Coherent Optics at 100G, 400G, and Beyond
Open RAN Platforms and Architectures Operator Survey
Cloud Native 5G Core Operator Survey
Bridging the Digital Divide
5G Network Slicing Operator Survey
Open, Automated & Programmable Transport
The Journey to Cloud Native
I don't think it was developed *because* of the 7950 - but Cisco has no
direct answer to that box on the market yet (Cisco would tell you to
count the CRS). That's what I meant.
What you're
saying about the LSR cards would make sense. I do think the CTR would
be a full-fledged router (it makes little sense for Cisco to not do that) with, as you said, an option to stuff it like an LSR, as the CRS platform has.
From what I've heard the approach will be like the CRS to have certain line cards available to do full routing, some will be more LSR functionality only. There should be 10x100G LSR cards available from the get go, but less density for the IPoDWDM type cards with fixed coherent/long-haul optics. Of course optics vendors are starting to come out with their own DWDM OTU4 optics to be used in a standard CFP slot.
The only thing I've heard I don't like is that it's a behemoth, and needs a full 23" rack to fit in.
I think the focus nominal port bit rates is misdirected. The major problem has been, and increasingly is, low revenue-generating traffic utilization of the theoretically network bandwidth rates. The reasons for the low effective utilization include all forms and layers of protocol stack and signaling etc overhead, as well as non-adaptive bandwidth physical layer channelization.
Is CTR delivering something to improve the bandwidth utilization rates, e.g. via the discussed multi-layer-optimization?
Overall, having an ASIC based hardware appears as a smart way to mitigate the potential SDN-over-whitebox hardware competition. But what do new CSCO ASICs do that generic (commercial chip based) hardware cannot, irrespective of how 'software-defined'?
Again, multi-layer-optimization in ASICs would seem like a clever approach.
Re: Full service router etc -- can't say for sure. My gut impression is that it is, but I don't have specifics to back that up.
It very much *will* start with 100G ports, just not 10 of them on a card all at once. I'd assume it'll start with 2x100G or 4x100G cards. The thing seems to be all about 100G density.
So, on the surface: Yes, I do think it's an answer to the 7950. Not saying it beats the 7950; you can think of the CTR as more a catchup if you want.