The industry really can’t stop talking about the edge. In the interest of not starting a classic “edge” conversation with a definition, let’s focus very specifically about its role as a new business model for 5G. From a Telco perspective, 5G enables a whole range of new applications and services that will open the doors to new revenue streams and provide a path to differentiation.
Aligning business outcomes to vertical industry requirements to operating models to technology enablers has become the foundation for how Dell Technologies is discussing the edge with our Telco customers and partners globally. For Telco, the “edge” is both a network architecture optimisation intended to improve experiences over cellular, but also a key control point for the development of differentiated services.
We talk about the edge as an enabler for 5G data-driven business models – real-time processing of data from “things” and devices, when the decisions need to be made quickly before the data loses relevance; but we also need to discuss how 5G is an enabler for the edge – bringing a disaggregated, cloud-native user-plane to allow applications and services to connect in closer proximity (and lower latency) to the “things” and devices. 5G and edge are symbiotes, even as both 5G and edge computing have use cases that are independent of each other.
While there are a lot of unknowns still as to what the edge will look like (Physically, obviously. Virtually, it will materialise as containers on bare metal), one thing for sure is there are no two identical edge deployments. Many questions need to be answered when designing the edge including: what are the use cases that are going to justify the investment — operations-first (driven by network function) or business-first (driven by enterprise use-case)? What are the unique vertical industry requirements? How impactful will the edge be to service experience? How significant will operational savings be?
Every situation is different –infrastructure at the edge is going to have to be very flexible to support a variety of use cases. The homogeneous infrastructure deployment model of public cloud – built with common building-blocks scaled within a rack, then within a row, then within the data centre – gives way to a heterogenous infrastructure model at the edge. A key to success is to have a flexible set of deployment capabilities that can be customised to support the specific edge requirements including:
- Ability to use multiple accelerators in a shared platform. This requires a heterogenous infrastructure with industry-standard servers and storage-class memory powered by x86 and accelerated with GPUs, FPGAs, or Smart NICs. This capability will enable edge deployments to meet the complex requirements to support video streaming or edge transcoding or any other functionality that drive various acceleration requirement at the edge.
- Infrastructure with flexible form factors to fit in a variety of environments, use-cases, and performance requirements. Infrastructure deployed outside a controlled environment needs to be hardened for harsh or rugged environments. It needs to be in a dense form factor and short-depth to optimise for space constraints. The telco network edge is still defined as front-facing, short-depth, DC-powered, NEBS-certified, but that’s just one of many telco deployment models.Aligning to the architectures of the 5G Cloud Native and Virtualised Network functions is key, such as converging user plane functionality from the vEPC with
- RAN Central Unit (CU) and potentially with the RAN Distributed Unit (DU). This helps to simplify the overall network architecture, reduce the cost of deployment and enhance the customer experience by centralising the 5G radios while distributing the vEPC functionality.
- Ability to easily perform remote management and automation, especially in lights-out, unmanned facilities. And to provide a set of APIs that allow that automation system to be integrated into service orchestration systems.
This week, Dell Technologies announced a portfolio of solutions that support edge environments. The new PowerEdge XE2420 is specifically designed for the above requirements. Namely, the new Dell EMC PowerEdge XE2420 server:
- Delivers performance and seamless scaling with a low-latency, high-performing short-depth system with high configuration flexibility.
- Supports up to four accelerators and 92 terabytes of storage, bringing dense compute and storage to space-constrained environments.
- Brings simplified remote management, streaming telemetry and automation capabilities via iDRAC9, including support for Redfish.
- Ensures reliability in the harshest edge environments including being able to operate in harsh, non-datacentre environments and more rugged Telco Network Edge requirements.
In addition to the PowerEdge XE2420, we’ve also refreshed the rugged PowerEdge XR2, a 1U, single-socket, shorter depth server that’s certified for telecom applications outside controlled environments. And we’re rolling out new Dell EMC Modular Data Centre Micro 415, that provides a hardened, rugged edge data centre with integrated power and cooling in an enclosure that fits in a parking space.
We also continue to expand our set of ecosystem partners who validate, certify, integrate with, and leverage Dell Technologies’ infrastructure capabilities in both the telco core and edge. But more on that later…