As we predicted in our 2012 report Cloud 2.0: Telco Strategies in the Cloud, operators have struggled to provide generically competitive cloud services, with those looking to provide infrastructure-as-a-service (IaaS) losing out to the larger hyperscale players (e.g. Amazon Web Services, Microsoft Azure). The majority of telcos have therefore reduced their focus and ambition within cloud (infrastructure) services over the last number of years.
However, recent legal and market developments and the emergence of new technologies are changing the cloud delivery model. The rescinding of the US-EU Safe Harbour agreement and the sovereign data trustee solution launched by Microsoft & Deutsche Telekom have put a spotlight on the need for sovereign cloud solutions that are better equipped to protect data. Operators are well-positioned to deliver and support these solutions but will need to act fast to ensure their role in the value chain.
Furthermore, new technologies (e.g. 5G, SDN/NFV) and requirements (e.g. low latency) may lead to the decentralisation of the current hyperscale data centre model, moving more computing power to the edge of the network (see How 5G is Disrupting Cloud and Network Strategy Today). This change in the architecture may lead to a long-term advantage for telcos.
In order to better understand data sovereignty requirements around the world and the potential opportunity for ‘sovereign’ cloud services, STL Partners (STL) conducted industry research. This research consisted of c.30 interviews with software-as-a-service (SaaS) providers, software companies, enterprises, public sector bodies, telecom operators and cloud service providers. This report presents and discusses the findings of this research.
The research programme was sponsored by Ericsson. This report and analysis was independently produced by STL Partners.
Introduction: The Return of Telco Cloud…
The telecoms industry has been undergoing a transformation process for much of the last decade. The threat from new players has marginalized the core communications business and operators have looked to gain traction and grow revenues through the provision of new services in adjacent areas, with one such area being cloud computing.
Cloud computing has ripped through the traditional IT infrastructure model, providing greater flexibility, enabling the pooling of resources and potentially reducing both capex and opex. This new delivery model has led to the development of new services and business models (e.g. ‘as-a-service’ models), disrupting how individuals consume services and how organisations do business.
The rise of cloud computing is a trend set to continue; indeed, STL Partners forecast that cloud IT infrastructure spending will equal spend on traditional IT infrastructure by 2020 (Figure 3).
Figure 3: Cloud IT infrastructure is rapidly gaining on traditional IT infrastructure
Source: IDC base figures; STL Partners analysis
Telcos have not remained oblivious to this industry transformation. Some (principally fixed-line) operators have a legacy providing IT outsourcing services and have looked to build on this footing, providing and managing infrastructure for cloud services, whilst others have partnered with cloud software providers to deliver new services to customers.
So far operators’ experiences offering cloud services have been mixed, with operators typically finding more success through partnerships. Rather than attempting to build their own cloud solutions operators have typically partnered with SaaS providers, such as Microsoft (e.g. Office 365) and Google (e.g. Google Apps for Work), acting as resellers of the software, potentially creating appealing bundles for enterprise customers.
On the other hand telcos attempting to provide IaaS, which one might intuitively think is more closely aligned to a telco’s core capabilities, have typically found that they have not been able to compete head-on with the larger IaaS providers (e.g. Amazon Web Services). Simply speaking, it has become a game of scale, with single operators or even telco groups unable to match the resources and investment of the hyperscale players. Indeed in our November 2014 report, Cloud: What is the role of telcos in cloud services in 2015?, we highlighted the challenge with telcos competing against the larger IaaS players:
“Pushing for pureplay IaaS solutions (Compute, Memory, Storage etc) is not going to be a sensible option for the majority of telcos. As an example of how hard it is to compete here, RackSpace came from a managed hosting/co location background and moved into IaaS, even collaborating on a virtualisation initiative that became OpenStack. But earlier in 2014, after spending less on IaaS investment than Microsoft or Google spend on infrastructure in a quarter, it announced it was going to refocus its efforts on its earlier product success with managed hosting and colocation because it was more able to differentiate itself from the other vendors who have significantly lower pricing.”
Telcos competing in infrastructure have therefore typically shifted their focus away from public cloud IaaS (competing against the larger providers) to more private cloud infrastructure and traditional managed hosting services. Despite mixed performance with IaaS services, albeit with exceptions in regions where the big IaaS players are not well established and where telcos can differentiate their offering (e.g. Telstra), there perhaps remains a still sizable opportunity, particularly as telcos begin to transform their networks.
This transformation involves the virtualisation of the network, embracing software defined-networking (SDN) and network functions virtualisation (NFV). As operators harness the power of these new technologies and associated business practices they will develop and implement the infrastructure, software and capabilities to deliver more advanced services through more efficient, automated and programmable networks. Operators in turn will be able to draw on these assets and associated skills to improve how they run and manage their cloud infrastructure.
Furthermore, as the industry develops and implements more advanced networks (i.e. 5G), there exists a potential advantage for telco infrastructure services due to the need for more localised delivery of service. The Next Generation Mobile Networks (NGMN) Alliance highlights that 5G should provide, “much greater throughput, much lower latency, ultra-high reliability, much higher connectivity density, and higher mobility range.”
STL Partners laid out a potential vision for 5G and network transformation in the report, How 5G is Disrupting Cloud and Network Strategy Today. To summarise the report, latency targets/requirements (how long it takes the network to respond to user requests) for 5G are very low; the target is 10ms end-to-end, 1ms for special use cases requiring low latency, or 50ms end-to-end for the “ultra-low cost broadband” use case. An example use case where low-latency could be very important could be communication between self-driving cars.
In order to meet these lofty requirements for latency the current delivery model may need to be rethought. Latency is limited by the time it takes to travel to the server and back at the speed of light; latency is therefore inherently linked to distance. In the 5G report, we explored the impact of these latency targets on the required distance of servers from users:
“The rule of thumb for speed-of-light delay is 4.9 microseconds for each kilometre of fibre with a refractive index of 1.47. 1ms – 1000 microseconds – equals about 204km in a straight line, assuming no routing delay. A response back is needed too, so divide that distance in half. As a result, in order to be compliant with the NGMN 5G requirements, all the network functions required to process a data call must be physically located within 100km, i.e. 1ms, of the user. And if the end-to-end requirement is taken seriously, the applications or content that they want must also be hosted within 1000km, i.e. 10ms, of the user. (In practice, there will be some delay contributed by serialisation, routing, and processing at the target server, so this would actually be somewhat more demanding.)”
To deliver these latency requirements a radical change to the architecture of the network is needed as well as a change in how compute and storage infrastructure is managed. Content and applications that are within the 100km contour will have a competitive advantage over those that don’t take account of latency. The impact of this could lead to the decentralisation of the current hyperscale data centre model, moving more computing power to the edge of the network. This change in the architecture and delivery model may lend telcos an advantage in the infrastructure marketplace.
Figure 4: Shifting the balance in favour of more localised infrastructure
Source: STL Partners
Whilst telcos will not wrestle control of the infrastructure marketplace overnight, telcos, as they embark on their transformation process, should look to make inroads towards this vision. Indeed there are current market challenges that telcos could immediately address (and are addressing) through their localised infrastructure, creating a stepped/phased approach towards the future vision of a localised cloud delivery model.
Into this rapidly evolving context steps the long-standing challenge of data sovereignty. Data sovereignty requirements are regulations that consider the implications of geographical location of data and place restrictions on the movement of certain types of data across borders. The recent ruling rescinding the US-EU Safe Harbour Agreement has put a spotlight on the issue of data privacy and data sovereignty and new approaches taken by technology players are highlighting that this is a problem that needs to and is being solved (i.e. Microsoft’s decision to create a German sovereign version of Azure). Operators are natural candidates to play a role here and should look to better understand how they can form part of the value chain in the provision of locally trusted IaaS solutions.
This report analyses data sovereignty requirements around the world and explores the potential opportunity for ‘sovereign’ cloud services as a further ‘nudge’ towards a more localised cloud delivery model.
- Executive Summary
- The Return of Telco Cloud…
- Understanding Data Sovereignty
- Which Sectors Have the Strongest Sovereignty Requirements?
- A Range of (Cloud) Solutions can Address Sovereignty Needs
- 75% of Interviewees were Interested in Sovereign Cloud Solutions
- Where is Data Sovereignty Important?
- How could this Evolve?
- Market Sizing: Sovereign Cloud could be Worth between $7-18bn in 2020
- Why Telcos are Well Positioned to Address the ‘Sovereign’ Opportunity
- Figure 1: A shift in the cloud delivery model may be occurring
- Figure 2: Sovereign cloud has the potential to represent over X% of the cloud infrastructure marketplace
- Figure 3: Cloud IT infrastructure is rapidly gaining on traditional IT infrastructure
- Figure 4: Shifting the balance in favour of more localised infrastructure
- Figure 5: How much data does Facebook store about you?
- Figure 6: STL Industry Research Programme – Breakdown of interviewees
- Figure 7: The significant majority of interviewees have encountered sovereignty requirements
- Figure 8: More-regulated sectors are more likely to encounter restrictions
- Figure 9: Infrastructure Deployment Models
- Figure 10: The applicability of cloud deployment models to meet sovereignty requirements
- Figure 11: The majority of Interviewees saw demand for sovereign cloud
- Figure 12: More strictly regulated sectors are more interested in sovereign cloud solutions
- Figure 13: Indicative map of data sovereignty requirements across the globe
- Figure 14: Overview of data sovereignty requirements across regions
- Figure 15: The rise of IoT could lead to increased demand for sovereign cloud
- Figure 16: Sovereign cloud could be worth between $7-18bn in 2020
- Figure 17: North America represents the biggest market for sovereign cloud
- Figure 18: Sovereign cloud in the Middle East & Africa potentially represents the greatest proportion of cloud infrastructure spending
- Figure 19: Government represents the largest market for sovereign cloud for existing services and Healthcare for sovereign cloud incl. IoT services
- Figure 20: Healthcare is the largest sector for sovereign cloud as a percentage of spend on IT infrastructure