Edge Looms Ever Larger; As Do Investment Requirements

The edge continues to move away from the edge, taking on more enterprise computing workloads and providing enhanced intelligence. However, supporting this movement will require substantial investments in new systems and hardware – thus making robust business cases even more critical to selling the concept to business leaders.

That’s the gist of a recent report out of the Linux Foundation, which looked at the impact of edge and 5G across key innovation areas such as critical infrastructure, hardware, networks, and software. The rise of edge is “increasingly viewed as a key enabler for the Fourth Industrial Revolution, in which the widespread deployment of the Internet of Things, the global sharing economy and the increase of zero marginal cost manufacturing deliver unprecedented communication-driven opportunities with massive economies of scale,” the report’s authors state.

Deployment of new edge infrastructure and applications expanded significantly over the past year, the Linux Foundation report states. Seven out of ten of the industry sectors explored saw increased forecasts in edge adoption over the previous year. To sustain this growth, sizeable infrastructure investments will be needed, the report states. “We estimate that between 2019 and 2028, cumulative capital expenditures of up to $800 billion will be spent on new and replacement IT server equipment and edge computing facilities — split between equipment for the device and infrastructure edges,” the report’s authors state.

There will also soon be a shift away from edge as an adjacent technology to cloud platforms to edge as a core computing environment in itself. In the short to medium term, “infrastructure edge demand for enterprise IT will be driven by cloud service use cases that are complemented and enhanced with edge computing capabilities,” the report notes. “However, in the long term, infrastructure edge demand will be driven by edge-native use cases that can only function when edge computing capabilities are available.” These use cases will involve technologies such as augmented and virtual reality as well as autonomous systems.

The report’s authors identified 43 use cases spanning 11 verticals, including CSP, enterprise IT, residential and mobile consumer services, retail, healthcare, automotive, commercial UAV, smart grid, smart cities, and manufacturing. Other verticals, such as education and financial services and investments at the device edge will drive additional edge computing market opportunities that are not included in the forecast.

There is also an extensive ecosystem of data centers and equipment now emerging to make edge viable for enterprises. “The foundational options for edge infrastructure are expanding, as hyperscale data center providers and telecom service providers jostle for position with suppliers of regional data centers, traditional colocation solutions, micro-edge data centers, wireless towers, and networking equipment.”

Interestingly, 5G technologies, while a significant advancement for edge systems, will not play a core role in the immediate future, the report’s authors claim. “While a segment of edge use cases will emerge that require 5G connectivity, 5G is not a requirement for mainstream edge applications. Many enterprise edge applications leverage wireline networks, while 4G-LTE and its variants LTE-M (Long Term Evolution category M1) and NB-IoT (Narrowband IoT) are adequate for the majority of current wireless connections. CBRS (Citizens Broadband Radio Service) and shared spectrum solutions are also appealing.”

The ”silicon” supporting many edge systems is also a key factor as such systems proliferate, shaping enterprise architectural decisions. Such decisions “become significant factors in CAPEX and OPEX as edge deployments scale-up and new silicon options emerge,” the report’s authors point out, adding that chip vendors and chip configurations are evolving. Intel, AMD, and Nvidia are in fierce competition, while acceleration options such as SmartNICs (Smart Network Interface Cards), FPGAs (Field-Programmable Gate Arrays), and DPUs (Data Processing Units) are aiming to boost the cost-performance of artificial intelligence and machine learning algorithms.