Industry Voices—Doyle: What hyperscale cloud and telecom partnerships could mean

One of the most interesting topics to me is the relationship between the hyperscale cloud providers, such as Amazon Web Services (AWS), Microsoft Azure and Google Cloud Project (GCP) with the equally large Tier 1 communications service providers (CSPs) or telcos. 

Both the IaaS providers and the CSPs have invested significantly in joint partnerships designed to deliver increased revenues. The cloud providers hope to sell their services into the large telecom vertical (approximately 10% of IT spend) and to gain market share against their competition. The telcos hope to leverage the efficiencies of cloud providers to reduce their infrastructure costs and to sell new 5G-based edge services.

This blog highlights five key findings related to Doyle Research's ongoing investigation into the critical intersection between cloud and telecom providers. What is clear is that despite the hype it is very early days for these partnerships. Revenue opportunities are highly speculative and business relationships are very fluid. Migration to the telco cloud – both edge and core – will be a long evolution, not a revolution.  And, Covid 19 impacts are pushing out time lines, which has led to delays in lab tests and early deployments.

Why cloud providers want to expand into the telecom vertical

The leading cloud providers are locked into a fierce battle to grow their revenues and to dominate the large (approaching $200 billion run rate) market for cloud services. AWS is the clear leader, followed by Azure and GCP remains a distant third—excluding the Chinese cloud providers for this discussion. IBM/Red Hat could be a factor in telecom, but has not announced specific plans or investments.  Other suppliers (potentially) in the mix include HPE, Oracle, and Dell.

Significant partnership announcements between cloud providers and CSPs include:

• Microsoft and AT&T

• AWS with Verizon, Vodafone, KDDI, and SK Telecom, among others.

Tier 1 CSPs are evaluating all the potential cloud providers and are likely to announce partnerships with two or all three leading cloud providers to explore the best opportunities. All partnerships are currently research projects with very little (if any) revenues. Business models remains very unclear – a critical point between these large service providers with competitive overlap and diverse agendas.

During my extensive time in the industry, I have seen very few (if any) partnerships between two large suppliers that meet expected revenue gains or have a significant market impact. The balance between cooperation and competition is very difficult to manage, especially in very large organizations.

CSPs want to improve the efficiencies of their networks AND grow revenues

Tier 1 CSPs want and need to change the economics of their business. They remain largely plumbing (bandwidth) suppliers with OK margins. Having missed out on the cloud and mobile application opportunities, CSPs are searching for new revenue sources (such as IoT and edge computing) that are enabled by increased optical bandwidth and 5G.

CSPs are also considering the use of core (IaaS) and edge cloud services to power part of their network operations. IT services will also be hosted in the cloud.  Some key potential use cases for hosting network applications in the cloud include:

• Network capacity surges

• Out of region coverage

• Back-up and reliability

• Edge computing and IoT service enablement.

Tier 1 telcos also want to leverage cloud partnerships to pressure their largest network suppliers (NEPs) into providing better (lower) pricing.   

Where are the cloud-native applications?

Many network applications – both core and edge – have migrated to virtualized software (NFV) running on x86 servers. Network ISVs are now working on containerized versions of their software, but with varying degrees of “cloud native." Very few network workloads are currently running in the IaaS cloud.  Potential applications include the following:

• BSS

• OSS

• Video—CDN

• Analytics

• Security

• Routing – driven by new software players such as DriveNets and Volta Networks

• RAN

None of these applications are a slam-dunk to migrate to the cloud due to various challenges of complexity, lack of cloud native functionality, cost, latency, etc.  Greenfield operators – e.g. Rakuten and Dish – are good candidates to place some of their network workloads in the cloud.

Cloud providers are acquiring telecom/networking talent

My LinkedIn notifications folder has been busy with announcements of experienced networking and telecom folks headed to GCP, AWS and Azure.  Microsoft has acquired two of the leading independent VNF/NFV vendors in Affirmed and Metaswitch for over $1 billion in total. Clearly, the cloud giants see opportunity in the network vertical. Many industry observers say it is not clear to the cloud providers how best to leverage the telecom opportunity.  They don’t necessarily benefit if they further hurt CSP margins and revenues, and they can be distracted by other large cloud opportunities if telco partnerships are slow to develop.

Edge computing is very nascent and business cases are unclear for telcos

Doyle Research believes the edge computing market will be slow to develop due to its diverse requirements, lack of standards and current lack of killer applications.  It is very expensive and time consuming to build out edge computing,  cloud giants have just started to build out their edge infrastructure.

RELATED: Doyle: 5 myths about multi-access edge computing

CSPs face the above challenges plus the following:

• Much of IoT connections will not be via 5G, Wi-Fi is more likely

• Telecom edge infrastructure is poorly designed to add edge computing (and RAN work loads are a completely separate issue)

• Partnership and complex revenue sharing models will be required to address the enterprise opportunities.

Edge computing may turn out to be a large market opportunity beyond 2025. It is likely that AWS, Azure and GCP will be well positioned here. Other suppliers are positioning their edge computing offerings include Equinix, Akamai, EdgeConneX, Saguna, Vapor IO, and Voltera – among many others.

Conclusions

The partnerships between cloud and telco providers are in a very nascent stage of development. We are in a period of discovery, trials and evaluation. For CSPs, working with leading cloud providers may offer efficiencies and cost savings for some network applications in some use cases. It will be several years before most network applications are truly cloud native.

New revenue opportunities will be difficult for leading CSPs. Both cloud and CSPs are trying to identify profitable business models for edge computing and for their respective partnerships. It will be interesting to see how these partnerships pan out over time—many different scenarios are possible and there are no guarantees of success.

Lee Doyle is Principal Analyst at Doyle Research, providing client focused targeted analysis on the Evolution of Intelligent Networks.  He has over 25 years’ experience analyzing the IT, network, and telecom markets. Doyle has written extensively on such topics as SDN, NFV, enterprise adoption of networking technologies, and IT-Telecom convergence. Before founding Doyle Research, Lee was Group VP for Network, Telecom, and Security research at IDC.  Lee holds a B.A. in Economics from Williams College. He can be reached at ldoyle@doyle-research.com and follow him @leedoyle_dc

Industry Voices are opinion columns written by outside contributors—often industry experts or analysts—who are invited to the conversation by Fierce staff. They do not represent the opinions of Fierce.