How the edge will look in 2020: predictions for the year ahead

Other predictions from Forrester, followed by our commentary in italics:

• “In 2020…unique demands of compute, storage, and networking for each use case will drive the development of custom form factors, right from the processor level.”

We generally agree on this point in regard to specialized chips being deployed for specific workload requirements. However, we note that the physical form needs to be standardized for easier installation and service for edge deployments, hence, efforts such as Open19. – Jim Davis, Edge Research Group

• “The edge cloud service market will grow by at least 50 percent.”

The public prediction seems to point to rapid growth of edge services from IaaS and PaaS services from hypercloud, telecoms and myriad other players. Given the number and size of the players, growing edge compute revenues from a relatively small revenue base makes this a fair target. – JD

• “Companies will choose multivendor packaged solutions over single vendors.”

Forrester suggests that the linkage of network, cloud, edge services and device functionality will require companies to work with vendors that have integrated functions specific to their industry.

Our research points to a similar conclusion. With the market for edge computing at an early stage, we believe vendors will have more success in forming a strong ecosystem of complementary technology and services by presenting complete solutions to customers. – JD

Rapid Edge Buildout paves way for more apps

Forrester has put a big stake in the ground with its edge computing coverage, but the other big research firms have some predictions they’ve laid out. IDC puts a stake its stake in the ground in terms of data center real estate.

IDC forecasts that more than 50 percent of new enterprise infrastructure deployed will be at the edge rather than inside the corporate data center by 2023. That compares to less than 10 percent at the start of 2020. All of the infrastructure will enable an 800 percent increase in the number of apps running at the edge by 2024.

Gartner’s take on the edge: More than half of enterprise data will be created and processed outside the data center or cloud. That processing will increasingly be done in place, on the edge. A consequence of this is that network design moves from a hierarchical architecture to a mesh architecture that is software defined and increasingly controlled by ML algorithms.

Data at the edge will be supported by a more distributed cloud, according to Gartner.

“Distributed cloud refers to the distribution of public cloud services to locations outside the cloud provider’s physical data centers, but which are still controlled by the provider. In distributed cloud, the cloud provider is responsible for all aspects of cloud service architecture, delivery, operations, governance and updates,” writes Gartner’s Kasey Penetta.

IoT growth, driven by edge device capabilities

Starting with foundational technologies, chips that enable faster processing of AI algorithms will experience exponential growth in 2020 and beyond, driven by consumer applications but augmented by enterprise uses. Sales of edge AI chips -which perform tasks machine learning on device-are expected to reach 750m devices for $2.6B in revenue in 2020, and reaching more than 1.5B units shipped by 2024, according to a forecast from consulting firm Deloitte. Smartphones will be the biggest driver, but smart home devices will be another factor in growth. In the enterprise market, edge AI chips will find a home in video surveillance and object recognition, for example.

Regardless of device type, one prediction that sadly will likely be too accurate: more IoT devices means increasing susceptibility to ransomware attacks in 2020, according to Forrester and many other researchers.

Security vendor RSA suggests that disruption on a wide scale will be a motivation for an attack on consumer devices, specifically personal assistants. The scale of the event could cause a widespread shift (downward) in consumer confidence in IoT endpoints

Perhaps widespread fear of that pessimistic prediction coming true will have a positive effect of increasing focus on security. Remember the Y2K bug at the turn of the century? Experts looking back on the bug say remediation efforts leading up to the year 2000 were instrumental in preventing disaster, though prognosticators of doom were proven wrong.

Article Topics

AI  |  chip  |  data center  |  edge computing  |  IoT  |  IoT security  |  ML  |  processor