Edge computing is the moving of IT infrastructure (servers, IT networking, and storage) closer to the end-user device or machine. This movement of IT infrastructure to locations closer to the point of its use is driven by the digitization of the enterprise and evolving expectations of consumers for applications that seamlessly integrate into real-time business processes and leisure activities. This has given rise to end-user applications that require very low latency response times and rapid processing of very large data sets. Examples include automation of manufacturing and logistics operations, enhanced security applications detecting suspicious behavior by monitoring video streams from public venues, AR-and-VR-driven applications in retail and immersive experiences in stadiums.
The movement of IT infrastructure closer to the end-user is contrary to the prevailing trend of moving application workloads to centralized, often cloud provider operated, data centers for scale-related cost reductions and ease of IT infrastructure management. In effect, the adoption of edge computing was warranted for data geography and sovereignty, and low latency coupled with the desire to maintain certain workloads in centralized data centers drives the development of a new IT ecosystem where edge computing use cases are realized by a coordinated set of workloads placed from cloud to the edge.
Adopting the notions put forward by community groups such as the Linux Foundation Networking group to use 20ms packet round trip time from an end-user device or a machine to demark the boundary between edge and centralized computation, it is clear that edge computing has been around for a long time: on-premises in the enterprise (shop floor and branch office) and points of presence of cloud service providers (Cloud SPs) and Telecom service providers (Telecom SPs).
In addition to the use of wireline connectivity for edge computing, the rollout of 5G and Multi-access Edge Computing (MEC) enhances the ability to provide low latency wireless connectivity between end-users and IT infrastructure, expanding the opportunity for Telecom SPs to participate in edge computing. MEC is covered in detail in Dell’Oro’s –Multi-Access Edge Computing Advanced Research Report. Edge computing has been made desirable by new use cases leveraging AI and ML technologies and cost-effective by new software development technologies such as micro-service architectures and distributed workload management platforms.
In this report, it is our objective to dig deeper into the IT ecosystem edge computing movement and assess the opportunity for equipment vendors, communication, and cloud providers.
The report addresses such questions as:
- What is edge computing, including a formal definition, involving the entire IT ecosystem?
- What are the top edge computing use cases?
- Who are the early adopters, by market segment, by region, and by enterprise verticals?
- What are the emerging edge computing solutions and disruptive vendors?
- How do AI and ML technologies get included in delivering edge computing use cases?
- How will the role of the Telecom SPs, Cloud SPs, and enterprises change with edge computing?
- What is the edge computing equipment opportunity over the next five years
- What IT technologies will need the largest investments over the next five years?
The report includes a 5-year forecast for the following areas:
- Edge computing IT equipment by location
- Ethernet Switching
- Data Center Power Infrastructure
- Edge computing IT equipment by customer segments
- Cloud SPs
- Telecom SPs
- Silicon for AI and ML, edge vs non-edge