2016 has seen a steady flow of announcements on successful Network Functions Virtualization (NFV) Proof-of-Concept deployments, mostly focused on virtualizing Customer Premise Equipment (vCPE). This has been a relatively straight forward starting point because unlike many other NFV applications, the vCPE use case does not involve complex activities like having to scale in or out individual services.
Yet there are only few service providers out there that would have taken NFV or SDN into full-scale production. Apparently, quite a few service providers are struggling to develop a business case that would justify large-scale rollouts. One of the reasons pointed out by cloud aficionados is that to fully embrace NFV, service providers would have to make significant CAPEX investments into new cloud data centres.
When NFV and SDN started making headlines, the sales pitch was based on CAPEX and OPEX reductions realized in regular service provisioning activities. This proposition failed to acknowledge that service providers are not starting their automation journey from a clean slate. Instead, many CSPs have their balance sheets full of CAPEX investments that will take years to digest.
At the same time, the incumbent network equipment manufacturers find the markets they serve shrinking in size. These dynamics lead to price erosion, which makes the traditional business case for NFV even more difficult. Against this backdrop, the savings that can be derived in the regular course of business are marginal at best.
Although no one really knows what the world will look in the next 10–15 years, there is a consensus that the number of connected devices will increase dramatically. Based on one recent prediction made by a staff member of the Singularity University, the number of connected devices would increase from 12 billion today to more than 100 billion Internet connected devices and a trillion sensors by 2025.
In the big scheme of things, the accuracy of predictions like this is not that important. Regardless of whether there will be hundreds of billions or trillions of connected devices, it is safe to argue that the monolithic networks of today will not be able to accommodate the traffic and the data that will flow through the networks. The largest problem lies in the on-demand tenet pervasive in cloud computing, because it requires completely new automated service architectures in order to deliver on its promise.
Therefore, instead of doing spreadsheet exercises focused on tinkering with the existing services and network infrastructure, service providers should take a more strategic view on the future. CAPEX and OPEX aside, the service provisioning industry is entering into a transformation whereby market shares will be redistributed to those who are able to adapt and to adjust.
On this note, there are two metrics that really count when preparing NFV and SDN business cases:
Cost of Scale represents the Total Cost of Ownership (TCO) associated with rolling out and operating environments that meet the scalability and the agility requirements of tomorrow. In this light, next-generation technologies such NFV and SDN will be able to offer significantly lower TCO than monolithic network technologies ever could.
Opportunity Cost represents the cost of foregoing the investments in next-generation technologies. While the opportunity cost today is not that significant, it is quite plausible that the market share, profitability and shareholder value delivered by the laggards will take a considerable hit over the coming years.
Considering the Darwian survival of the fittest that will shape the service provisioning industry over the next 10 years, it is a strategic imperative to begin the investments in automated service architectures today. The business case for these investments should be based on the Opportunity Cost of not doing anything, and the lower Cost of Scale provided by emerging technologies such as NFV and SDN.
By Juha Holkkola, Co-Founder and Chief Technologist at FusionLayer Inc.
|Data Center||Policy & Regulation|
|DNS Security||Regional Registries|
|Domain Names||Registry Services|
|Intellectual Property||Top-Level Domains|
|Internet of Things||Web|
|Internet Protocol||White Space|
Afilias - Mobile & Web Services
.eco launches globally at 16:00 UTC on April 25, 2017, when domains will be available on a first-come, first-serve basis. .eco is for businesses, non-profits and people committed to positive change for the planet. See list of registrars offering .eco more»