Communication service providers (CSPs) are active participants in delivering cloud-based applications and flexible, scalable networking services based on virtualized network infrastructures. A critical success factor in supporting these services is the efficiency and speed with which CSPs can create and deliver them to their customers. A widely accepted belief among CSPs about achieving this efficiency is that their future infrastructures should be designed using architectures based on the abstracted, software-defined, general-purpose infrastructure philosophies pioneered by hyper-scale cloud and application delivery companies. These would provide the elastic and adaptable foundation their future service offerings require.
Although this strategy holds great promise, CSPs have important choices to make in how they pursue it. One of those choices is in the type of hardware they employ in enabling these diverse, software-driven offerings. There are significant differences in how alternative hardware platforms are designed and how they fare in enabling the efficiency, versatility and scale CSPs require. Options range from using openly architected, standards-based solutions built for integration and versatility in a software-driven world; to solutions built using significant amounts of proprietary design and which are more challenging to integrate in an open platform environment; and platforms developed using a bare essentials white-box approach.
In its use of a standards-based, open architecture philosophy for its CSP portfolio, HPE has had a keen eye on the advantages this approach can deliver for the CSP. Still they need to be articulated and quantified. To accomplish this, HPE engaged with ACG Research to compare the total cost of ownership of the alternative approaches in CSPs’ deployments. This report describes the results of that analysis, which demonstrate significant economic advantages for CSPs when using the open architecture, standardsbased approach in their core and edge deployments for both NFV and cloudbased applications.