Mobile operators view multicloud as a competitive imperative, a strategy to avoid vendor lock-in, but running a 5G network on multiple clouds also creates additional complexity.

Other factors pushing operators to embrace a multicloud framework include customers’ cloud preferences, mobile edge computing capabilities, and specialized features available on different clouds. 

Myriad valuable outcomes are derived from multicloud, however it remains to be seen if wireless carriers will successfully blend these strengths in a cross-cloud fashion or operate them in effective silos. There are some, albeit not perfect, comparisons to be made to the radio access network (RAN) space where operators use equipment from multiple vendors but try to limit complexity by installing one vendor’s equipment in each market it serves.

Multicloud for operators today is a double-edged sword, according to Zeus Kerravala, founder and principal analyst at ZK Research. “Customers certainly want multicloud, but it does come with a lot of technical complexity and overhead,” he said.

“Operationally scaling in a multicloud scenario to meet the needs and demands of a telco environment has an impact on speed and time to market for an operator to deploy its services,” Kerravala said. 

As such, operators need to weigh costs versus the technical and supplier-specific benefits when deploying cloud-based workloads or services, he explained. 

The multicloud model is to “create a service architecture that brings to bear a capability that increases multiple cloud services to the customer but also internally in terms of how it migrates its internal network and IP stack to the cloud,” Nick McQuire, chief of enterprise research at CCS Insight, explained in a recent phone interview.

Operators are also effectively hedging their cloud investments by partnering with multiple clouds, and taking a methodical approach to vet the cloud providers — project by project, workload by workload, McQuire said. 

Aspirations and reality aren’t fully aligned, though. “Because of the complexity and overhead, most telcos will start with a single cloud provider, most likely Amazon Web Services (AWS) and then look to move to multicloud later,” Kerravala said.

Nonetheless, “it’s important that communications service providers do not hitch their wagon to a single cloud provider just yet,” IDC Senior Analyst Patrick Filkins said, adding that every hyperscaler is gaining traction in private networks and edge cloud infrastructure. 

“At the end of the day, the enterprise will have a lot to say about which cloud provider will run which workloads. If a communications service provider is only attached to one cloud provider, that may limit their market potential,” he said. 

Relatively few operators have gone all-in with one cloud provider, but there are outliers like Dish Network, which is housing everything it can in the AWS cloud, including network workloads, RAN compute, and various forms of edge computing.

Verizon, which previously intended to use AWS for public edge cloud and Microsoft Azure for private edge cloud, recently signaled a growing preference for AWS as its public and private edge cloud provider. AWS is also the only commercially available option for Verizon today.

“It may be OK to run internal workloads on a single cloud provider, in fact that may end up being an optimal choice, however, the communications service provider will likely need to have a multicloud play, especially if a client wants their app running on a specific cloud or edge cloud platform,” Filkins said.

Patrick Kelly, founder and principal analyst at Appledore Research, is also dubious about exclusive arrangements between operators and cloud providers. “Exclusive cloud deals don’t make sense because each hyperscaler brings with it strong value propositions based on business outcomes,” he said.