Network Cloud Transformation: Lessons from the Enterprise Experience for CSPs

For years, enterprise leaders have successfully transformed siloed, vertically stacked IT environments into cloud operating models that converge compute, storage, and networking into a single pool of resources.

MeriTalk recently sat down with Dell Technologies’ León Taiman, Client Principal, Global Telco Practice, to discuss the parallels between enterprise IT and telecom network transformations and lessons learned that directly apply to Communications Service Providers as they plan their future networks.

MeriTalk: Take us back a few years and talk about the early days of enterprise cloud transformation and the role Dell played.

Taiman: It’s been an amazing journey as we saw enterprise clients through three phases of IT transformation over the past decade and a half.

The first phase emerged with the adoption of virtualization technologies. It started with going from physical to virtual environments, where customers could replace physical silos within which applications were vertically integrated with and tightly coupled to purpose-built hardware with virtual machines. Virtualization broke down those silos, enabling customers to run workloads on common hardware, bringing greater agility as well as efficiencies in capex and connecting customers to a storage area network.

The second phase involved the adoption of self service, automation and cloud infrastructure. Some customers wanted the benefits of cloud but did not have the expertise or time to transform their IT environments. As a result, they turned to the public cloud, initially thinking that their cost and operating complexity would go down. In time, they learned that for many applications the public cloud costs more. This phase also introduced concerns about security, data sovereignty, and a siloed operating model. Other customers built their own private cloud, many of them in partnership with Dell Technologies, and reaped the benefits of lowered costs, improved agility, and a more secure environment.

Finally, the third phase involves a move towards a cloud-native architecture that is multi-cloud by design. Customers are re-factoring and migrating their application workloads to be container based with Kubernetes emerging as the standard. With a cloud native based private cloud, the resiliency is built into the application rather than the infrastructure, and automation through Infrastructure as Code is embedded, dramatically improving performance for the application owner and developer teams. This phase involved the most dramatic shift in our customers’ operating models.

Dell Technologies has been working hands-on with our customers through all three of these phases delivering compute, networking, and storage infrastructure, supported by cloud platform partnerships to deliver an open, multi-cloud by design cloud platform that enables a cloud operating model spanning on-premises and public cloud environments seamlessly. And that really drives agility and a significant reduction in the total cost of operations (TCO).

MeriTalk: What parallels do you see between the enterprise IT transformation and the telco network transformation process?

Taiman: The biggest parallel between enterprise transformation and where telco networks are today is that both had a siloed based operating model, with enterprises adopting a cloud operating model sooner. Both enterprises and telcos are highly dependent on their application providers, however enterprises were able to pressure their enterprise application providers to refactor and certify on a common cloud native platform sooner. Telcos continue to depend on their network equipment providers (NEPs) to deliver the full stack end-to-end vertical solution. As the IT enterprises moved from a vertical siloed physical to virtual to cloud to cloud-native, the telco NEPs kept offering a full stack, vertical environment. The delay in adopting virtualization and then cloud-native technologies occurred because of the risk adverse nature of the telecom industry and the misperception that these application workloads would be unable to run reliably in these modern cloud native environments.

Over the last two to three years, NEPs have realized that if they do not embrace open cloud native platforms, disaggregation, and horizontal cloud platforms they are going to fall behind. In many cases, Tier 1 operators are telling the NEPs that they need to run on a horizontal cloud with open APIs or they will not be considered as vendors going forward. This gives the telcos an opportunity to leapfrog some of the technology hurdles their enterprise peers went through, and jump to multi-cloud by design, cloud-native microservice-based architecture.

We are now seeing a 40 to 60 percent reduction in capex and opex as they transform from a virtual to a cloud-native environment, with greater agility to leverage all the data across those silos, which is critical because data is the fuel for AI, just as we learned in enterprise. These TCO reduction opportunities have already been achieved by many enterprises that started this transformation several years ago.

MeriTalk: What are some of the best practices you recommend for communications service providers (CSPs)?

Taiman: I have always advised CSPs to start with one telco domain and to ‘walk before you run’ and get one domain done right before extending transformation to other domains.

The decision to migrate to a horizontal cloud needs to come from the top-down as it is a business-case driven decision, and in that process, we help our customers quantify reductions in opex and capex and increased business agility.

It is key that CSPs also have clear alignment with the NEPs in this transformational journey towards a telco cloud.

I also emphasize the importance of how future-state cloud operating models are essential, as this type of transformation requires re-tooling and changes in people and processes. Without these models, CSPs run the risk of deploying great technology but not being able to operationalize it.

MeriTalk: As CSPs think about managing edge and AI workloads, what do they need to know about storage?

Taiman: Data is the fuel for AI, and the development of AI use cases in the data management and storage environment is critical. A challenge for the telcos today is that a lot of their data is spread across many databases and storage environments that are siloed by nature in legacy environments both on premises and across several cloud providers. Both Enterprises and Telco’s have this common challenge, which is why Dell Technologies has developed an open, virtual, multi-cloud Data Lakehouse solution to facilitate the consumption of data to drive AI workloads. In addition, Software Define Storage technologies across Block, File and Object storage offer the ability to virtualize storage and enable seamless data mobility with secured access.

Moving data across silos or to the public cloud does not work for AI if the goal is to monetize. Where we help customers is bringing AI to their data and not the other way around. We advise CSPs to consolidate and create that virtual data lake house with software-defined storage infrastructure spanning file, block, and object storage so you are fueling your AI models without moving your data around. This enables them to virtualize the data and present it to their AI models in a consistent fashion.

When you transform your network, you are collapsing compute, storage, and networking into a pooled resource that relies on APIs and can be dynamically applied to changing business needs and managed easily throughout the lifecycle.

In enterprise IT transformation, storage virtualization lagged compute and delayed achieving the benefits of adopting a cloud operating model. Telco’s can avoid this mistake and drive towards a target cloud platform that incorporates virtualization across compute, network, and storage infrastructure. This will allow storage to be viewed as workload agnostic, where storage capacity, performance and cost can be optimized depending on the use case.

Unlike enterprise, a CSP’s storage differs in scale and scope, accounting for far edge all the way back to the national data centers. CSPs need software-defined storage that is reachable throughout their network, like banks with branches and ATMs living at the far edges of the network.

Enterprise IT transformation highlights the significant opportunity that exists for telcos to become more agile – to gain the ability to leverage their data across silos and use that data for emerging AI models and use cases that improve operations and power new services. Because of the experience with enterprise IT, we know CSPs are well positioned to accelerate their cloud transformation strategies.