Virtualisation business cases reveal the real NFV deal for CSPs
As communications service providers (CSPs) engage with virtualisation technologies in general and network functions virtualisation in particular, it’s clear that chief financial officers need a more detailed understanding of specific virtualisation business cases in order to approve the required investment. To gain this rigorous financial information, Amdocs has commissioned analyst firm Analysys Mason to develop a series of NFV business cases. Last year, the research focused on virtualised customer premise equipment (vCPE) and the firm was able to construct clear business cases for CSPs to follow. Now, the companies are working to construct business cases that focus on the impact NFV has on video.
Preliminary findings from the latest research into video use cases were shared at TMF Live! 2016 in Nice, France on 9-12 May.
The first of the new business cases looks at the virtualisation of fixed video or content delivery networks (CDN) and the second looks at the business case for virtualisation of mobile networks to support video and mobile edge computing (MEC). Here Larry Goldman, a partner and the head of telecoms software research at Analysys Mason, and Justin Paul, the head of OSS marketing at Amdocs, discuss their findings to date with VanillaPlus.
VanillaPlus: Tell us about the new research Analysys Mason has been involved in around NFV business cases?
Larry Goldman: We’ve conducted research into how the NFV business is developing and how CSPs are starting to deploy services based on NFV for several years. Last year, Amdocs came to us because it wanted to better understand the business cases for NFV and to be able to explain in financial terms the benefits of adopting NFV. For CSPs, it comes down to having a good, solid business case to enable them to determine whether certain projects will have a pay off.
Justin Paul: CSPs get the concept of NFV in engineering terms but now they want to make significant investments in the technology, they need a business case. Some major tier-1 CSPs are setting aggressive targets for adoption and they wouldn’t be going so rapidly if there wasn’t a business case but others are struggling more with understanding the business case. We’re working to provide those and demonstrate the really good business reasons that exist off the back of this technology. There are very strong business reasons to adopt SDN and NFV and we are working with Analysys Mason to showcase those.
VP: Why are NFV business cases important to the industry?
LG: As an industry, people look at areas that it seems NFV will benefit and some CSPs are pushing forward with deployments based on the fact that NFV in general seems to make sense. It’s clear that the shift from costly, function-specific boxes to commodity IT hardware will generate some savings and that the move to software has value. CSPs can see that from the surface and from a simple, capex-oriented point of view NFV makes sense.
However, to deploy NFV you have to put resources in place and you have to be able to explain to the CFO in CFO terms what the net present value of an investment is and how to get to a good, legitimate underlying model. The industry is moving on from saying this seems to make sense and lots of CSPs are waiting to move forward. They understand NFV’s benefits conceptually but want to understand what it’s really worth. Someone’s got to do the math.
VP: The businesses cases are both focused on video, why is video getting so much focus in the industry right now?
JP: We’ve selected video use cases as our next area of research because the impact of video is so significant to CSPs. With the emergence of 4k and 8k video and the completely different video consumption habits of generation Z, the challenges CSPs face are becoming clear. They need increased capacity but also agility to cater for peaks in demand. They therefore need the ability to store content where it’s easily accessible and put it in a centralised, distributed locations.
The issues with video are going to be the same throughout the world and we’re aware we need a rigorous bottom up cost base analysis.
LG: Our first two business cases focused on vCPE in the enterprise and in the residential markets. Those both have substantial tangible costs in terms of installation and truck roll that virtualisation can readily minimise and therefore business cases are simpler to construct. We’ve selected video this time because it’s an area of high investment for CSPs because there’s so much growth in demand that they have to address.
CSPs know they have to make this investment so they want to know if virtualisation makes sense in this area.
VP: What were the preliminary results of your survey?
LG: I shared the findings in greater detail in my presentation at TMF Live! 2016 in Nice, France but in the video case there is a capex saving for both video platforms and CDNs. The CDN savings are notably better than for the video platform.
One interesting variation we have uncovered is that while virtual functions can be hosted in either private cloud or public cloud, the public cloud makes sense for very small CSPs but the public cloud charges for storage and transit make them cost prohibitive for large CSPs.
VP: What did you learn about virtual probes (vProbes) in the fixed CDN business case? How important were they to making the business case positive?
JP: The fixed video network model uses virtual probes instead of physical probes. This is because traditional, physical probes can’t probe a virtual network and the cost of a virtual probe is significantly lower than a physical one. We’re working with Radcom to implement a vProbe solution with a North American CSP and we’re seeing from the work we’ve done there that physical probe is 20-25% less costly than a physical probe. In addition, you can throw up a ring of probes around a specific area to address a specific peak in demand and redeploy those licences elsewhere when the peak has passed. They’re cheaper to buy and they offer greater flexibility and agility to operators because of that redeployment capability.
LG: Virtual probes collect key data that is important for two factors that influence the business case. One is around understanding customer usage patterns the other is around optimising network traffic. The virtual probe aspect of NFV deployment is turning into a very interesting area in its own right because it offers substantial capex and opex reductions for CSPs.
VP: If service providers look at the figures in your business case, but don’t think they reflect their home market how can they get a business case that reflects their cost and customer base?
JP: Analysys Mason will be looking at how much it costs to send an engineer to a site and how much truck roll costs. It’s clear the costs in Europe and North America for this will be similar but in Sao Paolo or India the cost base for physical activities will be different and will have an impact on the business case. We’re hoping to get a much broader segmentation of CSP markets covered in this research and, while we will put out a generic business case for each use case, we will use the data we collect to build specific use cases for customers on request.