Summary Bullets:
- Among others, AT&T and DT have both committed to commercial service launches that will run on a virtualized network architecture
- If two of the most aggressive rollouts in the world are only virtualizing fairly rote services, is that a bad sign for near-term vendor sales prospects?
So, anyone who heard AT&T’s John Donovan speak at Mobile World Congress had to come away impressed with the scope and aggressiveness of the carrier’s progress on Domain 2.0. Among the more noteworthy highlights were:
- Retraining 130,000 employees (We’ve been hearing operational change is a huge deal, right?)
- Plans to virtualize 75% of its functions by 2020 (per AT&T that equates to 150 unique network functions)
- The company is already offering services that run on virtualized infrastructure (Dang, that was fast)
If you were at MWC and pay attention to network infrastructure, you probably also got wind of DT and the launch of its pan-European cloud-based VPN service, running over a mix of virtualized and physical network elements. This, too, is quite impressive in the speed at which the carrier has been able to put SDN/NFV to work, so to speak.
Against that backdrop, would it be in poor taste to ask, “Is that it?”
I mean, Ethernet services? VPNs? At their core, both are fairly basic connectivity services. And, according to both carriers, the traffic traversing these virtualized networks will all be “net new” traffic for the time being. To be clear, I’ve got no problem with that. It is the prudent course. Why run existing traffic over an architecture that has barely been invented yet, let alone just being rolled out?
However, a question that should be asked is, “What does it mean for the near-term sales prospects for equipment vendors?” If two of the most aggressive SDN/NFV rollouts are still fairly cautious with respect to what they trust to SDN/NFV, are we in for what in the past has been an inevitable trough in momentum behind new technology roll-outs?
My guess? Yes… but not for a while yet.
My guess is that 2015 will be a gangbusters type year from revenue growth associated with sales of SDN and/or NFV products and services. We’re already seeing awards from operator on a global basis. Telstra, Telefonica, Telecom Italia, NTT, and many others were linked to product awards in the past few months alone. It’s no bold statement to say that these operators all share the vision of using virtualization to carry net-new traffic. Doing so helps cut the recurring cost of supporting new services, and helps capture new revenue, all while risking relatively little in the way of an existing revenue base. There’s a reason “cap and grow” is a term, after all.
That said, this is the easy part. Cutting over live existing traffic, especially traffic that is sensitive to latency or requires very strict security standards, will be a much slower burn. This is when you might expect to see a slow-down in SDN/NFV sales happening. When will this be?
I’m not sure. But, if I was taking a survey, I’d say 24-36 months is a good bet…for now.