Network edge capacity forecast: The role of hyperscalers

Developers need to see sufficient edge capacity

Edge computing comprises a spectrum of potential locations and technologies designed to bring processing power closer to the end-device and source of data, outside of a central data centre or cloud. This report focuses on forecasting capacity at the network edge – i.e. edge computing at edge data centres owned (and usually operated) by telecoms operators. 

This forecast models capacity at these sites for non-RAN workloads. In other words, processing for enterprise or consumer applications and the distributed core network functions required to support them. We cover forecasts on RAN as part of our Telco Cloud research services portfolio.

Forecast scope in terms of edge locations and workload types

Source: STL Partners

Enter your details below to request an extract of the report

The output of the forecast focuses on capacity: number of edge data centres and servers

STL Partners has always argued that for network edge to take off, developers and enterprises need to see sufficient edge capacity to transform their applications to leverage its benefits at scale. The forecast seeks to provide an indication for how this will grow over the next five years, by predicting the number of edge data centres owned by telecoms operators and how many servers they plan to fill these up with.

Hardware vendors have been evolving their server portfolios for a number of years to fit the needs of the telecoms industry. This started with core network virtualisation, as the industry moved away from an appliance-based model to using common-off-the-shelf hardware to support the virtualised LTE core.

As infrastructure moves “deeper” into the edge, the requirements for servers will change. Servers at RAN base stations will not have full data centre structures, but need to be self-contained and ruggedised. 

However, at this stage of the market’s maturity, most servers at the network edge will be in data centre-like facilities. 

There are three key factors determining a telco’s approach and timing for its edge computing data centres

Telecoms operators want to build their network edge capacity where there is demand. In general, the approach has been to create a deployment strategy for network edge data centres that guarantees a level of (low) latency for a certain level of population coverage. In interviews with operators, this has often ranged from 90-99% of the population experiencing sub-10 to 20 millisecond roundtrip latency for applications hosted at their network edge.

The resultant distribution of edge capacity will therefore be impacted by the spread of the population, the size of the country and the telecoms operator’s network topology. For example, in well connected, small countries, such as the Netherlands, low latencies are already achievable with the current networks and location of centralised data centres.

Key factors determining network edge build​

Source: STL Partners

The actual number of sites and speed at which a telecoms operator deploys these sites is driven by three main factors: 

Factor 1: edge computing strategy;

Factor 2: the speed at which it has or will deploy 5G (if it is a mobile operator);

Factor 3: the country’s geographic profile.

Details on the evidence for the individual factors can be found in the inaugural report, Forecasting capacity of network edge computing.

Table of contents

  • Executive summary
  • Introduction to the forecast
  • Key findings this year
  • Regional deep-dives
  • Role of hyperscalers
  • Conclusions
  • Appendix: Methodology

Enter your details below to request an extract of the report

Telco cloud: short-term pain, long-term gain

Telcos have invested in telco cloud for several years: Where’s the RoI?

Over a number of years – starting in around 2014, and gathering pace from 2016 onwards – telcos have invested a large amount of money and effort on the development and deployment of their ‘telco cloud’ infrastructure, virtualised network functions (VNFs), and associated operations: long enough to expect to see measurable returns. As we set out later in this report, operators initially hoped that virtualisation would make their networks cheaper to run, or at least that it would prevent the cost of scaling up their networks to meet surging demand from spiralling out of control. The assumption was that buying commercial off-the-shelf (COTS) hardware and running network functions as software over it would work out less costly than buying proprietary network appliances from the vendors. Therefore, all things being equal, virtualisation should have translated into lower opex and capex.

However, when scrutinising operators’ reported financials over the past six years, it is impossible to determine whether this has been the case or not:

  • First, the goalposts are constantly shifting in the telecoms world, especially in recent years when massive 5G and fibre roll-outs have translated into substantial capex increases for many operators. But this does not mean that what they buy is more (or less) expensive per unit, just that they need more of it.
  • Most virtualisation effort has gone into core networks, which do not represent a large proportion of an operator’s cost base. In fact, overall expenditure on the core is dwarfed by what needs to be spent on the fixed and mobile access networks. As a ballpark estimate, for example, the Radio Access Network (RAN) represents 60% of mobile network capex.
  • Finally, most large telco groups are integrated operators that report capex or opex (or both) for their fixed and mobile units as a whole; this makes it even more difficult to identify any cost savings related to mobile core or any other virtualisation.

For this reason, when STL Partners set out to assess the economic benefit of virtualisation in the first half of 2022, it quickly became apparent that the only way to do this would be through talking directly to telcos’ CTOs and principal network engineers, and to those selling virtualisation solutions to them. Accordingly, STL Partners carried out an intensive interview programme among leading operators and vendors to find out how they quantify the benefits, financial or otherwise, from telco cloud.

What emerged was a complex and nuanced picture: while telcos struggle to demonstrate RoI from their network cloudification activities to date, many other benefits have accrued, and telcos are growing in their conviction that further cloudification is essential to meet the business, innovation and technology challenges that lie ahead – many of which cannot (yet) be quantified.

The people we spoke to comprised senior, programme-leading engineers, executives and strategists from eight operators and five vendors.

The operators concerned included: four Tier-1 players, three Tier-2 and one Tier-3. These telcos were also evenly split across the three deployment pathways explained below: two Pathway 1 (single-vendor/full-stack); three Pathway 2 (vendor-supported best-of-breed); and three Pathway 3 (DIY best-of-breed).

Four of the vendors interviewed were leading global providers of telco cloud platforms, infrastructure and integration services, and one was a challenger vendor focused on the 5G Standalone (SA) core. The figure below represents the geographical distribution of our interviewees, both telcos and vendors. Although we lacked interviewees from the APAC region and did not gain access to any Chinese operators, we were able to gain some regional insight through interviewing a new entrant in one of the major Asian markets.

Geographical distribution of STL Partners’ telco cloud benefit survey

 

Source: STL Partners

Virtualisation will go through three phases, corresponding to three deployment pathways

This process of telco cloudification has already gone through two phases and is entering a third phase, as illustrated below and as decribed in our Telco Cloud Manifesto, published in March 2021:

Phases of telco cloudification

Source: STL Partners

Effectively, each of these phases represents an approximately three to five-year investment cycle. Telcos have begun these investments at different times: Tier-1 telcos are generally now in the midst of their Phase 2 investments. By contrast, Tier-2s and -3s, smaller MNOs, and Tier-1s in developing markets are generally still going through their initial, Phase 1 investments in virtualisation.

Given that the leading Tier-1 players are now well into their second virtualisation investment cycle, it seems reasonable to expect that they would be able to demonstrate a return on investment from the first phase. This is particularly apt in that telcos entered into the first phase – Network Functions Virtualisation (NFV) – with the specific goal of achieving quantifiable financial and operational benefits, such as:

  • Reduction in operational and capital expenditures (opex and capex), resulting from the ability to deliver and run NFs from software running on COTS hardware (cheaper per unit, but also more likely to attract economies of scale), rather than from expensive, dedicated equipment requiring ongoing, vendor-provided support, maintenance and upgrades
  • Greater scalability and resource efficiency, resulting from the ability to dynamically increase or decrease the capacity of network-function Virtual Machines (VMs), or to create new instances of them to meet fluctuating network capacity and throughput requirements, rather than having to purchase and maintain over-specified, redundant physical appliances and facilities to guarantee the same sort of capacity and resilience
  • Generation of new revenue streams, resulting from the ability that the software-centricity of virtualised networks provides to rapidly innovate and activate services that more closely address customer needs.

Problem: With a few exceptions, telcos cannot demonstrate RoI from virtualisation

Some of the leading telco advocates of virtualisation have claimed variously to have achieved capex and/or opex reductions, and increases in top-line revenues, thanks to their telco cloud investments. For example, in January 2022, it was reported that some technical modelling had vindicated the cost-reduction claims of Japanese greenfield, ‘cloud-native’ operator Rakuten Mobile: it showed that Rakuten’s capex per cell site was around 40% lower, and its opex 30% lower, than the MNO incumbents in the same market. Some of the savings derived from automation gains related to virtualisation, allowing cell sites to be activated and run remotely on practically a ‘plug and play’ basis.

Similarly, Vodafone claimed in 2020 that it had reduced the cost of its mobile cores by 50% by running them as VNFs on the VMware telco cloud platform.

The problem is that the few telcos that are willing to quantify the success of their virtualisation programmes in this way are those that have championed telco cloud most vocally. And these telcos have also gone further and deeper with cloudification than the greater mass of the industry, and are now pushing on with Phase 3 virtualisation: full cloud-native. This means that they are under a greater pressure to lay claim to positive RoI and are able to muster data points of different types that appear to demonstrate real benefits, without being explicit about the baseline underpinning their claims: what their costs and revenues would, or might, have been had they persisted with the old physical appliance-centric model.

But this is an unreal comparison. Virtualisation has arisen because telco networks need to do more, and different things, than the old appliance-dependent networks enabled them to do. In the colourful expression of one of the industry experts we interviewed as part of our research, this is like comparing a horse to a computer.

In the first part of this report, we discuss the reasons why telcos generally cannot unequivocally demonstrate RoI from their telco cloud investments to date. In the second part, we discuss the range of benefits, actual and prospective, that telcos and vendors have observed from network cloudification, broken down by the three main pathways that telcos are following, as referred to above.

Enter your details below to request an extract of the report

Table of Contents

  • Executive Summary
  • Telcos have invested in telco cloud for several years: Where’s the RoI?
    • Virtualisation will go through three phases, corresponding to three deployment pathways
    • Problem: With a few exceptions, telcos cannot demonstrate RoI from virtualisation
  • Why do operators struggle to demonstrate RoI from their telco cloud investments to date?
    • For some players, it is clear that NFV did not generate RoI
    • It has also proved impossible to measure any gains, even if achieved
  • Is virtualisation so important that RoI does not matter?
  • Short-term pain for long-term gain: Why telco cloud is mission-critical
    • Cost savings are achievable
    • Operational efficiencies also gather pace as telcos progress through the telco cloud phases
    • Virtualisation both drives and is driven by organisational and process change
    • Cloud-native and CI/CD are restructuring telcos’ business models and cost base
  • Conclusion: Telco cloud benefits are deferred but assured
  • Index

Related research

Driving the agility flywheel: the stepwise journey to agile

Agility is front of mind, now more than ever

Telecoms operators today face an increasingly challenging market, with pressure coming from new non-telco competitors, the demands of unfamiliar B2B2X business models that emerge from new enterprise opportunities across industries and the need to make significant investments in 5G. As the telecoms industry undergoes these changes, operators are considering how best to realise commercial opportunities, particularly in enterprise markets, through new types of value-added services and capabilities that 5G can bring.

However, operators need to be able to react to not just near-term known opportunities as they arise but ready themselves for opportunities that are still being imagined. With such uncertainty, agility, with the quick responsiveness and unified focus it implies, is integral to an operator’s continued success and its ability to capitalise on these opportunities.

Traditional linear supply models are now being complemented by more interconnected ecosystems of customers and partners. Innovation of products and services is a primary function of these decentralised supply models. Ecosystems allow the disparate needs of participants to be met through highly configurable assets rather than waiting for a centralised player to understand the complete picture. This emphasises the importance of programmability in maximising the value returned on your assets, both in end-to-end solutions you deliver, and in those where you are providing a component of another party’s system. The need for agility has never been stronger, and this has accelerated transformation initiatives within operators in recent years.

Enter your details below to request an extract of the report

Concepts of agility have crystallised in meaning

In 2015, STL Partners published a report on ‘The Agile Operator: 5 key ways to meet the agility challenge’, exploring the concept and characteristics of operator agility, including what it means to operators, key areas of agility and the challenges in the agile transformation. Today, the definition of agility remains as broad as in 2015 but many concepts of agility have crystallised through wider acceptance of the importance of the construct across different parts of the organisation.

Agility today is a pervasive philosophy of incremental innovation learned from software development that emphasises both speed of innovation at scale and carrier-grade resilience. This is achieved through cloud native modular architectures and practices such as sprints, DevOps and continuous integration and continuous delivery (CI/CD) – occurring in virtuous cycle we call the agility flywheel.

The Agility Flywheel

agility-flywheel

Source: STL Partners

Six years ago, operators were largely looking to borrow only certain elements of cloud native for adoption in specific pockets within the organisation, such as IT. Now, the cloud model is more widely embraced across the business and telcos profess ambitions to become software-centric companies.

Same problem, different constraints

Cloud native is the most fundamental version of the componentised cloud software vision and progress towards this ideal of agility has been heavily constrained by operators’ underlying capabilities. In 2015, operators were just starting to embark on their network virtualisation journeys with barriers such as siloed legacy IT stacks, inelastic infrastructures and software lifecycles that were architecture constrained. Though these barriers continue to be a challenge for many, the operators at the forefront – now unhindered by these basic constraints – have been driving a resurgence and general acceleration towards agility organisation-wide, facing new challenges around the unknowns underpinning the requirements of future capabilities.

With 5G, the network itself is designed as cloud native from the ground up, as are the leading edge of enterprise applications recently deployed by operators, alleviating by design some of the constraints on operators’ ability to become more agile. Uncertainty around what future opportunities will look like and how to support them requires agility to run deep into all of an operators’ processes and capabilities. Though there is a vast raft of other opportunities that do not need cloud native, ultimately the market is evolving in this direction and operators should benchmark ambitions on the leading edge, with a plan to get there incrementally. This report looks to address the following key question:

Given the flexibility and driving force that 5G provides, how can operators take advantage of recent enablers to drive greater agility and thrive in the current pace of change?

Enter your details below to request an extract of the report

 

Table of Contents

    • Executive Summary
    • Agility is front of mind, now more than ever
      • Concepts of agility have crystallised in meaning
      • Same problem, different constraints
    • Ambitions to be a software-centric business
      • Cloudification is supporting the need for agility
      • A balance between seemingly opposing concepts
    • You are only as agile as your slowest limb
      • Agility is achieved stepwise across three fronts
      • Agile IT and networks in the decoupled model
      • Renewed need for orchestration that is dynamic
      • Enabling and monetising telco capabilities
      • Creating momentum for the agility flywheel
    • Recommendations and conclusions

The future of assurance: How to deliver quality of service at the edge

Why does edge assurance matter?

The assurance of telecoms networks is one of the most important application areas for analytics, automation and AI (A3) across telcos operations. In a previous report estimating the potential value of A3 across telcos’ core business, including networks, customer channels, sales and marketing, we estimated that service assurance accounts for nearly 10% of the total potential value of A3 (see the report A3 for telcos: Mapping the financial value). The only area of greater combined value was in resource management across telecoms existing networks and planned deployments.

Within service assurance, the biggest value buckets are self-healing networks, impact on customer experience and churn, and dynamic SLA management. This estimate was developed through a bottom up analysis of specific applications for automation, analytics and AI within each segment, and their potential to deliver cost savings or revenue uplift for an average sized telecoms operator (see the original report for the full methodology).

Breakdown of the value of A3 in service assurance, US$ millions

Breakdown of the value of A3 in service assurance (US$ millions)

Source: STL Partners, Charlotte Patrick Consult

While this previous research demonstrates there is significant value for telcos in improving assurance on their legacy networks, over the next five years edge assurance will become an increasingly important topic for operators.

What we mean by edge assurance is the new capabilities operators will require to enable visibility across much more distributed, cloud-based networks, and monitoring of a wider and more dynamic range of services and devices, in order to deliver high quality experience and self-healing networks. This need is driven by operators’ accelerating adoption of virtualisation and software-defined networking, for example with increasing experimentation and excitement around open RAN, as well as some operators’ ambitions to play a significant role in the edge computing market (see our report Telco edge computing: How to partner with hyperscalers for analysis of telcos’ ambitions in edge computing).

To give an idea of the scale of the challenge ahead of operators in assuring increasingly distributed network functions and infrastructure, STL Partners’ expects a Tier-1 operator will deploy more than 8,000 edge servers to support virtual RAN by 2025 (see Building telco edge infrastructure: MEC, private LTE and vRAN for the full forecasts).

Forecast of Tier 1 operator edge servers by domain

Forecast of Tier-1 operator edge servers by domain

Source: STL Partners

Given this dramatic shift in network operations, without new edge assurance capabilities:

  • A telco will not be able to understand where issues are occurring across the (virtualised) network and the underlying infrastructure, and diagnose the root cause
  • The promises of cost saving and better customer experience from self-healing networks will not be fully realised in next-generation networks
  • Potential revenue generators such as network slicing and URLLC will be of limited value to customers if the telco can’t offer sufficient SLAs on reliability, latency and visibility
  • It will not be possible to make promises to ecosystem partners around service quality.

Despite the significant number of unknowns in the future of telco activities around 5G, IoT and edge computing, this research ventures a framework to allow telcos to plan for their future service assurance needs. The first section describes the drivers affecting telcos decision-making around the types of assurance that they need at the edge. The second sets out products and capabilities that will be required and types of assurance products that telcos could create and monetise.

Enter your details below to request an extract of the report

Table of contents

  • Executive Summary
    • The three main telco strategies in edge assurance
    • What exactly do telcos need to assure?
  • Why edge assurance matters
  • Factors affecting edge assurance development
    • What are telcos measuring?
    • Internal assurance applications
    • Location of measurement and analysis
    • Ownership status of equipment and assets being assured
    • Requirements of external assurance users
    • Requirements from specific applications
    • Telco business model
  • The status of edge assurance and recommendations for telcos
    • Edge assurance vendors
    • Telco assurance products
  • Appendix

Enter your details below to request an extract of the report