Of the more than 700 IT decision makers surveyed for the report by 451 Research, 56 percent say at least a quarter of their data centre and cloud capacity is already supporting IoT.
By Martin T. Olsen, VP of Global Edge and Integrated Solutions, Vertiv.
That represents an enormous amount of capacity for an application that hasn’t yet achieved widespread deployment. Only 33 percent of survey participants have broadly deployed internet of things, with an additional 26 percent in limited deployment and 38 percent in some stage of planning.
As those limited deployments and plans become broad deployments, the percent of IT capacity supporting IoT will increase. Eighty-nine percent of companies surveyed expect at least a quarter of their capacity to be supporting IoT in two years, with almost half expecting at least 50 percent of capacity to be supporting IoT.
Increasing IT Workloads
IoT is likely the major reason a large majority of those surveyed expect their IT workloads to grow in the next two years. And, the massive volumes of data that will be generated is forcing a dramatic shift away from on-premises data storage. While almost three-quarters of enterprises surveyed currently store IoT data on-premises, that is expected to drop to just 27 percent in a year. That data will move to cloud hosting and colocation data centres as well as to network operator infrastructure and devices and systems at the edge.
Reason For Defining Edge Archetypes
As our work on defining edge archetypes makes clear, many edge applications, while relying on cloud computing for data archiving and deep learning, require distributed IT capacity that is located close to the source of data.
The Vertiv report Defining Four Edge Archetypes and their Technology Requirements, identifies four primary edge archetypes that encompass the leading use cases for edge computing. These archetypes are Data Intensive, Human-Latency Sensitive, Machine-to-Machine Latency Sensitive, and Life Critical.
IoT applications, such as smart buildings, cities and factories, fall within the Data Intensive Archetype because the volume of data generated makes it unfeasible or cost prohibitive to rely exclusively on the cloud for data processing. This can create the need for a “fog” layer between the device and the cloud. As the Data at the Edge report states:
Fog nodes, which are computing facilities that sit between IoT data generators and the centralised cloud, can include micro-modular datacentres, as well as larger datacentres located near to end devices (things), including colocation and other leased facilities, and intelligent IoT gateways. Fog nodes can handle IoT workloads generating massive data volumes that are inefficient to transport to a centralised location, as well as IoT workloads that require low latency.
Growing IoT Data – What Is The Advantage For Colocation Providers?
So, it’s clear that internet of things data volumes are growing, moving off-premises and, in many cases, requiring processing close to the point of generation. As a result of those factors, colocation providers with buildouts of physically distributed points of presence have an advantage over centralised cloud providers in capitalising on the need for IoT capacity.
That advantage could be strengthened by the move to 5G wireless technologies. The massive data volume generated by the billions of remote and mobile IoT devices and sensors creates the potential for current cell tower sites to evolve to support a distributed colo model operated by current or emerging colocation providers.
The opportunity exists today for forward-thinking colos to begin targeting and capturing demand for IT capacity driven by IoT deployments. According to the Data at the Edge report, “Special attention should be focused on those verticals and countries with the highest proportion of enterprises in the late planning stages for IoT support. These prospects are likely to be considering the capacity impacts of IoT and therefore will be interested in data storage and processing options.”
Which Enterprises Plan To Manage The Growth Of IoT?
Interestingly, those verticals and countries tend to be the ones that currently have a low percentage of enterprises that have already broadly deployed IoT.
For example, China has a high percentage of enterprises that have broadly deployed IoT (55 percent) but a small percent (nine percent) in the late planning stage. By contrast, the United Kingdom and Italy are below 20 percent in broad deployments but are among the countries with the highest percentage of companies in the late planning stages, making them fertile territory for colocation providers positioned to support fog computing.
Vertical markets most likely to be in the late planning stage include government and higher education.