As anyone who's met me knows, it's not often that I'm stuck for words, but the huge changes that are currently sweeping the telecom industry this year have had even me struggling for the appropriate vocabulary to describe what's going on.
Fortunately, on a recent trip to Silicon Valley I heard a new industry term doing the rounds -- one that I think encompasses the current telecom zeitgeist perfectly.
The term is "the New IP" (and yes, it implies that there is also an "Old IP," of which we shall speak more shortly).
So what is the New IP? In simple technology terms, it's a state-of-the-art, virtualized IP network. But the reason why service providers and large enterprises need to sit up and take notice of the New IP doesn't have to do with technology nearly as much as it has to do with money.
It's about the money, dummy!
Specifically, networks built on the New IP technology help service providers to both save money on capex (capital expenditure) and opex (operational expenditure), and make money by selling new content-driven services.
This sets them apart from networks built on the Old IP, which typically don't do either of those things, consisting, for the most part, of utilitarian pipes designed to reliably deliver mucho traffic from hither to whence (and whose most important characteristic as far as service providers go is that they cost money).
Ok, so I'm simplifying things here -- but the telecom industry is actually well overdue for a dose of simplification, isn't it? The reason service providers today are anxious about having to plan their future network strategies is that the tech media (with the exception of Light Reading, obviously) has done an absolutely splendid job of obscuring the importance of virtualization by smearing it in a thick and bewildering layer of technical jargon, acronyms, buzzwords, and hype.
Sure, everyone has heard about SDN and NFV (boy, haven't we?!), but those specifications aren't what matters here; it's how they revolutionize service providers' businesses that counts (note to telcos everywhere -- "follow the money," in other words). These acronyms are actually symptoms of an absolutely huge upheaval that's about to turn the telecom ecosystem (and its economics) on its head.
Old, new, or somewhere in between
I'll dig into how the New IP changes everything in a minute. But first, gentle Light Reader, let's work out if YOUR telecom network is part of the Old IP world, or part of the New IP nirvana.
To do that, Light Reading has devised the following simple quiz to show how far you've come in moving from old to new, and how that progress could affect your company's bottom line.
The rules are simple: give yourself one point (1 pt) for each New IP box that your network checks, subtract one point (-1 pt) from the total for each box you check in the Old IP column, and then compare your score to our handy "What Your Score Means" key.
Table 1: Are You More New IP Than Old IP?
|The Old IP||The New IP|
|Designed to scale clients (devices/nodes)||Capable of scaling clients and resources on-demand (cloud-like)|
|Rigid topology and architecture||Fluid in topology and architecture|
|Integrated control and data planes||Disaggregated control and data planes|
|Decentralized intelligence and management||Centralized intelligence and management|
|Proprietary but standards-driven innovation||Open platform and open-sourced innovation|
|Time-bound provisioning and change management||On-demand provisioning and programmability|
|Key success metric: performance (speeds and feeds)||Key success metric: agility (usability)|
|Killer apps: data networking communications (email), ecommerce, voice/video/data integration (VoIP, unified comms)||Killer apps: Cloud everything, mobile data centers, big data analytics, virtualization everywhere|
|Management considers your network as an essential budget line item||Management considers your network to be a strategic asset that serves both the bottom and the top line by saving/making money|
What your score means
- +10 points (max. possible)
Whoa! Check out the big brain on [YOUR NAME HERE]. Are you sure you don't come from the future? Because, dude, your network like totally defines state of the art. Would you like a job at Light Reading? Scratch that -- can we all come and work for you?
- +9 to +6 points
All hail, IP visionary, we salute you! Where other service providers enervate, you innovate. Truly you are a trail blazer on the information superhighway, charting a course that the rest of the telecom industry must follow.
- +5 points to +1 point
Congratulations, slugger -- you've clearly done battle with the latest in comms technology and come out a winner. Your network isn't bleeding edge, but you've made real strides towards dragging it into the 21st Century.
- 0 points to -9 points
Humph. The good news is that your IP network is no worse than 90% of those currently installed by service providers. The bad? You are paying too much to buy and run a network which is actually making it harder for you to deploy new and profitable applications and services. Ouchy!
- -10 points (worst possible score)
It's pretty dark where you are, isn't it, old timer?
The shock of the new
As mentioned, one reason service providers are getting pumped up about virtualization is that it has the potential to deliver a double whammy of benefits: saving money on capex and opex AND making money on new services.
On the capex side, virtualization reduces the number of router ports that service providers have to buy -- especially out at the edge of the network. Conversely, core routers aren't much affected, and optical gear not at all -- or, as Michael Capuano, VP of marketing at Infinera Corp. (Nasdaq: INFN), told me the other day, "as far as I'm aware, no one has yet invented a way to create photons in software."
Note: It's a clear indicator of the robust health of the telecom market that Tier 1 service providers such as AT&T Inc. (NYSE: T) and Verizon Communications Inc. (NYSE: VZ) aren't planning to use those edge router savings to reduce their net capex spend (which, when combined, added up to an astonishing $21 billion in 2012). Instead, they're expected to continue spending at record levels for at least the next year or so. What virtualization means is that this spend will purchase a lot more infrastructure for service providers' money -- good for them and their customers (and the rest of the telecom supply chain, of course).
The capex savings are nice, but the opex benefits delivered by the New IP are potentially even more significant than the capex savings.
Those opex savings come in different varieties. Use of automation to enable on-demand services reduces the need for humans to be involved in service provisioning, thus decreasing the inevitable incidence of human error on telecom networks. At the network level, service providers can reduce the number of truck rolls required to provision new services by doing it with a mouse instead.
But this new opex model isn't just about saving, it's also about enabling. In fact, what the New IP enables is probably even more important than any cost savings. At the business development level, the New IP gives the telcos the ability to start acting much more like their new OTT competitors (Netflix, Amazon, Facebook and so on) by using their live networks as a venue to design, spec, test and deploy new and profitable content-driven digital services. By building on open-sourced standards, network operators should also be able to act as better partners to innovators and developers, and more quickly incorporate outside innovation into their service mix.
And there's more. A New IP environment hands greater control and flexibility to a service provider's customers, who will be able (policies and rules permitting, of course) to add/drop and flex the services and applications they're using without having to reach out to their service provider and wait for a response/action. All of that adds up to an improved customer experience and an environment that encourages experimentation with, and use of, a greater number of services.
All this represents a meta-shift from where service providers are today: The switch from Old IP to New IP is as big, in fact, as the change from circuit-based to packet-switched networks.
Having written about tech for 25 years I've seen more than my share of technology hype, but having spent this year talking to service providers and their equipment suppliers I'm convinced that the New IP is the real deal.
In fact, the only thing likely to prevent service providers from taking New IP technology and running with it is the service providers themselves. Culturally, while the telcos want to compete with the new wave of OTT competitors such as Amazon, and recognize that in order to do so they must act more like the OTT players, in day-to-day reality most of them are miles away from being able to do so. Proprietary attitudes and "worst practice" for customer relations that go back literally decades will have to be overcome before real change can take place.
In other words: the ability of the telcos to make full use of the New IP will have as much to do with their own business philosophies as it does with technology.
Next page: All change?