Edge computing is a growing technology field that exploits computing at local and edge locations before typically connecting and communicating with the cloud. Often this has to do with the sheer volume of data generated at the edge and which would not be feasible or sensible to send to the cloud when it can be first processed at source before sending select data to the cloud. The setting up of the Linux Foundation Edge (LF Edge) by the Linux Foundation in January 2019, follows the success of its Cloud Native Computing Foundation (CNCF) and has already attracted more than 60 premier and general-level members and onboarded several open source projects. LF Edge aims to create an open source edge framework independent of hardware, cloud, and OS. Ovum is seeing the vendor community ramp up their edge computing activity as the rise of five technologies and factors are converging on the edge: 5G, artificial intelligence (AI), cloud-native computing technologies (microservices, containerization, and DevOps), increased compute power available in the edge, and the reduction of the cost of computing at the edge.
What the edge constitutes will vary across players. For a large public cloud provider, the edge might be a data center, whereas for a car manufacturer, the edge could be a vehicle. LF Edge says the edge is defined by devices that are local, possibly mobile but must have low-latency responses in less than 20ms. The mobile phone is excluded by LF Edge from its scope by having an established technology stack, but the definition works well for a broad range of edge cases.
The convergence of technologies is helping make edge computing possible, including greater computational power at the edge; fast, high-bandwidth communications with low latency through 5G and NFV; AI allowing sophisticated decision-making and processing at the data source; microservices and containerization with DevOps-style continuous delivery simplifying edge application development and deployment; powerful low-power microprocessors and accelerators (especially for AI applications); and the lowering of compute costs. Together these make edge computing possible and desirable and will lead to innovative IoT edge applications.
Existing Linux Foundation projects Akraino Edge Stack, EdgeX Foundry, and Open Glossary of Edge Computing have been moved into LF Edge, and new projects have been created. Project EVE contributed by Zededa provides an edge virtualization container technology, and Home Edge Project by Samsung Electronics provides a real-time data hub for home devices. The aim is to create a common middle layer through which any edge device can communicate with the cloud. This means edge devices can focus on data computation without having to reinvent a proprietary wheel to communicate with the cloud or other devices.
Michael Azoff, Distinguished Analyst, Infrastructure Solutions